Chapter 1 — Definition, Landscape and Stakes

On a gray morning in Sherbrooke, Karim Haddad printed the second-year results of Hydrosentinelle inc. and laid them beside a sheet the two cofounders had filled in twenty-six months earlier, when the company had no customers at all. The two sheets agreed on exactly two lines and disagreed on almost every other. The old sheet said that an installed kit of sensors and a gateway would sell for $4,800 and cost $2,100, and that a monitoring subscription would sell for $79 a month and cost $15 a month to serve. The new sheet said that the company had billed $891,000 in its second year against a plan of $910,200, had earned $110,400 of earnings before interest, taxes, depreciation and amortization, and was entering its third year with 206 buildings under subscription. The two lines that agreed were the two prices, and nobody had decided to keep them; and neither cofounder could say, in a sentence a customer would accept, why the kit cost what it cost. This chapter is about the decision the cofounders had not yet made, and about what it is worth. It defines a price and separates it from the words owners use interchangeably with it; it describes how prices actually get set in businesses of this size; it proves, on this company's own third year, that price is the strongest profit lever it has; it names the challenges that keep the lever untouched, the needs a pricing decision must satisfy, the constraints that bound it, and the problems a price cannot solve; and it says how the rest of the book is organized. Book A1 (From Idea to Business Model), Chapter 7, forms one defensible price hypothesis per segment at the discovery stage and stops there, on purpose. This book takes over after that hypothesis exists.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Define a price precisely, and separate it from cost, value, revenue, margin and pocket price, which are five different things that owners routinely treat as one.
• Recognize how your own price was actually set, among the six patterns that account for most prices in small businesses, and name the symptom each pattern produces.
• Compute the four profit levers of your business on one year of your own figures, and say in dollars what one percentage point on each is worth.
• Name the six things a business needs from its pricing, and say which of them your business currently has in writing.
• Name the five constraints that bound a price — law, competition, capacity, cost and data — and say which one is binding for you right now.
• Say what pricing is not responsible for, so that a targeting problem, an offer problem or a sales problem is not treated as a pricing problem.

Why It Matters

A price is the only number in a business that converts directly into profit. Every other improvement passes through something else: more units must be made and delivered, a cheaper supplier must be found and qualified, a fixed cost must be renegotiated or removed. A price that goes up by one percent, and that customers still pay, adds its full amount to the bottom line on the day it takes effect. It matters not because price increases are easy — they are not, and Chapter 8 treats one as a serious operation — but because most owners spend their scarce attention on the weakest lever they have, since the weakest lever feels the most controllable.

For a solo founder, the stakes are survival and self-respect in roughly equal measure. A solo business that prices from its costs is letting its suppliers and its own inexperience set its income, and it usually discovers the mistake only when it is fully booked and still not making a living. For a small team, the stakes are consistency: the moment two people can quote, the business has either a policy or a lottery, and a lottery produces a spread of prices for the same work that nobody can explain to a customer who compares notes with another customer. For an established business, the stakes are compounding. A price set once and left alone falls every year in real terms while every cost rises, and a discount granted once becomes the reference price of that account forever, so the margin lost is not lost for one deal but for the life of the relationship.

There is also a legal dimension that no other marketing decision carries so directly. A price is a representation made to the public every time it is displayed, and in Canada a false or misleading price representation is reviewable conduct under the Competition Act and, when made knowingly or recklessly, a criminal offense. Section 1.6 sets out that boundary and Chapters 3, 6 and 7 return to it provision by provision. Pricing is the place in a business where a marketing decision becomes a legal question immediately, and an owner who does not know that boundary will eventually cross it while trying to be clever.

In This Chapter

• 1.1 What a Price Is, and What It Is Not
• 1.2 The Landscape: How Prices Actually Get Set in Small Businesses
• 1.3 Why Price Is the Strongest Profit Lever
• 1.4 The Challenges: Cost Anchoring, Fear and the Absence of a Policy
• 1.5 What a Business Needs from Its Pricing
• 1.6 The Constraints: Law, Competition, Capacity, Cost and Data
• 1.7 What Pricing Is Not Responsible For
• 1.8 How This Book Is Organized

1.1 What a Price Is, and What It Is Not

1.1.1 A price is a proposed exchange, not a property of a product

A price is the amount of money a seller asks in exchange for a defined thing, delivered on defined terms, to a defined buyer, at a defined moment. Most of the words in that sentence do work. The amount is what the seller asks, not what the thing is worth and not what it cost; asking is a decision, and it can be made differently tomorrow. The terms matter as much as the amount, because thirty days of credit, a deposit, a warranty and an installation included or excluded all move the real price without moving the number. The buyer matters, because the same product legitimately carries different prices for buyers who buy differently.

What follows is the single most useful idea in this book: a price is not a property of a product in the way that its weight or its color is. It is the output of a decision, made by a person, on evidence that can be gathered or ignored. A business that speaks of "our price" as though it were discovered rather than chosen has already given up the decision. Hydrosentinelle's $4,800 kit price is not a fact about the kit; it is a decision two people made in a week, on the evidence they had then, and never revisited.

1.1.2 Five words that are not the price

Owners use five other words as though they were synonyms for price, and every one of them describes something different. Cost is what the business gives up to produce and deliver the thing: for Hydrosentinelle, $2,100 for an installed kit and $15 a month to monitor a building. Cost is determined by suppliers, by process and by scale, and it is the one number in the list that the customer neither sees nor cares about. Value is what the thing is worth to the buyer, measured against what the buyer would otherwise do; it is determined by the buyer's situation, and it is usually much larger than the price and entirely invisible from inside the seller's office. Revenue is price multiplied by quantity, which is why revenue can rise while the business gets worse. Margin is price minus cost, in dollars or as a percentage of price; Hydrosentinelle's kit carries $2,700 of margin, or 56.25 percent, and its subscription carries $64 a month, or 81.01 percent. And pocket price is what the business actually keeps from a transaction after every discount, allowance, freight concession, extended payment term and unbilled overrun has been subtracted — the only one of the five that is measured after the fact rather than decided in advance, and the one that most businesses have never computed. Chapter 6 computes it for one Hydrosentinelle deal and finds that 30.59 percent of the list margin had disappeared across six lines, of which not one had been negotiated as a discount.

DEFINITION Price, cost, value, revenue, margin and pocket price A price is the amount of money a seller asks for a defined offer, on defined terms, from a defined buyer, at a defined moment; it is a decision, not a property of the thing sold. A cost is what the business gives up to produce and deliver that offer, determined by suppliers, process and scale rather than by the market. Value is what the offer is worth to the buyer relative to the buyer's best alternative, including doing nothing; it is the ceiling a price can approach and never a number the seller controls. Revenue is price multiplied by quantity sold, which is why it can grow while the business becomes less profitable. Margin is price minus cost, in dollars as unit margin or as a percentage of price as margin rate; it is not the same as markup, which expresses the same gap as a percentage of cost, and Chapter 2 shows what confusing the two costs. Pocket price is what the business actually keeps once every discount, allowance, credit, freight concession, extended payment term and unbilled overrun has been taken out of the invoice price; the gap between list price and pocket price is where most small businesses lose margin without noticing. In plain terms: cost is what you give up, value is what the buyer gains, price is what you ask, revenue is what you collect, margin is what is left, and pocket price is what is really left.

1.1.3 The relation between the three that matter

Cost, competition and value stand in a fixed relation to each other, and the relation is the foundation of every method in Chapter 2. Cost sets a floor: below it, the business is paying customers to take its product, which is a defensible thing to do only deliberately and briefly. Competition sets a reference: the buyer will compare, and the comparison will be made whether or not the seller participates in it. Value sets a ceiling: no rational buyer pays more than the offer is worth relative to the alternative. The strategic pricing literature has argued for decades that the ordering matters: value is established first, and cost is used to decide whether the business should be in the market at all, not to compute the price (Nagle et al., 2023). Most small businesses reverse the order, and Section 1.4 explains why.

The following table, Table 1.1, sets the six words side by side so that a reader can check which of them a colleague actually means in a conversation about "the price." Its columns give the word, its meaning in this book, what determines it, and the thing it cannot tell you on its own. The column to read most carefully is the last one, because almost every pricing mistake in this book comes from asking one of these six numbers a question that only another one can answer.

Table 1.1 — Six words owners use as if they meant the same thing

1.1.4 What a price communicates, whether or not you intend it

A price is read by buyers as information about the thing being sold, and that reading happens before any argument the seller makes. Book B1 (Marketing Fundamentals for Business Owners), Chapter 6, states the three readings plainly: a price well below the alternatives says "cheaper, and probably less"; a price well above them says "better, and here is why"; and a hidden price says "expensive, and embarrassed," which sends the buyer to a competitor who displays one. B1 owns the price as a signal inside the marketing mix, and this book owns the methods that produce the number the signal carries; Chapter 7 returns to whether a price should be published at all.

1.2 The Landscape: How Prices Actually Get Set in Small Businesses

1.2.1 What the national data say about prices in motion

Canadian data do not measure how businesses choose a price, but they measure how often a price moves, and that is nearly as informative. In its analysis of small businesses for the second quarter of 2026, Statistics Canada reports that 24.3 percent of businesses with 1 to 19 employees expected to increase the selling price of the goods and services they offer over the following three months, while 65.2 percent expected prices to stay the same and 7.1 percent expected them to decrease; among businesses with 20 to 99 employees the share expecting an increase was 31.2 percent, and among those with 100 or more employees it was 27.4 percent (Statistics Canada, 2026a). The same analysis reports that 64.0 percent of businesses with 1 to 19 employees expected at least one cost-related obstacle over the same three months, with 48.7 percent naming rising inflation and 27.1 percent naming rising input costs.

Put those two measurements next to each other and the landscape appears. Roughly two-thirds of the smallest businesses in the country expected their costs to be an obstacle, and roughly two-thirds of them expected their prices to stand still. That is not a forecast of stability; it is a forecast of margin compression, quarter after quarter, in the businesses least able to absorb it.

1.2.2 The six ways a price actually gets set

In practice, the price a small business charges arrives by one of six routes, and it is worth being able to name the route your own price took. The first is cost plus a habitual markup, where a number the owner learned somewhere is applied to a cost that may itself be incomplete. The second is matching a visible competitor, usually a single competitor whose price is easy to find and whose cost structure is unknown. The third is the number that felt sayable, arrived at by imagining the moment of telling a customer and stopping at the largest figure that did not produce a flinch. The fourth is inherited from the first customer, where a price quoted once became the company's price forever. The fifth is set by the largest buyer, where a major account negotiated terms early and every later quote was anchored to them. The sixth is never set at all: the price has not changed since launch and no one remembers deciding it.

None of the six is illegitimate as a starting point, and Hydrosentinelle's own prices came from the first and the third combined. The problem is not the origin of the price; it is that the origin is never examined, so the business cannot say what would have to change for the price to change. That is the difference between having a price and having pricing.

The following table, Table 1.2, sets out the six routes with what an owner typically says about each, what in fact determined the number, and the symptom the route produces later. Its columns run from the route to the symptom, and the symptom column lets a reader work backwards from what is going wrong today to how the price was set.

Table 1.2 — How a price actually gets set in a small business, and what each route produces

1.2.3 Who actually decides, and who actually gives it away

There is a second landscape behind the first, and it concerns people rather than numbers. In most small businesses the person who sets the price and the person who gives it away are not the same person, and neither of them has been told that pricing is part of the job. The owner sets a list price once; the person in front of the customer then adjusts it, deal by deal, under pressure, without a floor, without an approval step and without any record that the adjustment happened. The list price is therefore a fiction after the first quarter, and the real price distribution of the business is unknown to everyone in it.

CASE IN POINT Hydrosentinelle enters year 3 with a price set in year 1 Fictional example; names and figures are invented. Hydrosentinelle inc. sells an installed kit of connected water-leak sensors and a gateway to owners and managers of commercial and rental buildings, with a monthly monitoring subscription attached. The kit has been $4,800 since the first sale, against a cost of $2,100; the subscription has been $79 a month against a cost of $15; and an annual prepayment has always carried a 10 percent discount, for $853.20 instead of twelve monthly payments. In year 2 the company billed $891,000 against a plan of $910,200, a shortfall of $19,200, and earned $110,400 before interest, taxes, depreciation and amortization. It enters year 3 with 206 buildings under subscription and a plan to install 180 kits. Laurence Bergeron, the president, believes the kit is underpriced and has said so twice without evidence; Karim Haddad believes any increase will cost volume the company cannot afford to lose, and has said so twice without evidence either. Neither cofounder has computed what a price change would be worth, what it would cost, or what the company's own customers already reveal about what they will pay. The whole of this book is the work those two cofounders had not yet done.

1.3 Why Price Is the Strongest Profit Lever

1.3.1 The arithmetic, in four lines

A business has four levers on its operating profit, and only four. It can raise its price, sell more volume, reduce its variable cost per unit, or reduce its fixed operating costs. The four are not equal, for a reason that is arithmetic rather than opinion.

A one percent price improvement, with volume unchanged, adds one percent of revenue to profit, because no additional cost is incurred to collect it. A one percent volume improvement adds one percent of gross margin, because the additional units must be produced and delivered before they contribute. A one percent reduction in variable cost adds one percent of total variable cost. A one percent reduction in fixed costs adds one percent of operating expenses. The observation that the right price moves profit faster than additional volume does is old and well established in the management literature; Marn and Rosiello put it at the center of their 1992 study of pricing as a profit discipline, arguing that getting the price right is the fastest and most effective way for a company to reach its maximum profit (Marn & Rosiello, 1992).

1.3.2 The four levers on the case, measured

Hydrosentinelle's third year gives the arithmetic a body. The plan for the year is 180 kits at $4,800, which is $864,000, and an average of 281 subscriptions at $79 a month, which is $266,388, for total revenue of $1,130,388. Variable costs are $428,580, leaving a gross margin of $701,808, a gross margin rate of 62.09 percent. Operating costs are $520,000, which leaves $181,808 of operating profit.

DO THE MATH The four profit levers compared on one year Fictional example; figures are invented and computed by the case model. Year 3 of Hydrosentinelle inc. plans revenue of $1,130,388, variable costs of $428,580, a gross margin of $701,808 at 62.09 percent, operating costs of $520,000 and operating profit of $181,808. Price: one percent on every price, with volume unchanged, is one percent of revenue, or $11,303.88, which is 6.22 percent of the operating profit. Volume: one percent more units sold is one percent of gross margin, or $7,018.08, which is 3.86 percent. Fixed costs: one percent off operating costs is $5,200.00, which is 2.86 percent. Variable cost: one percent off the cost of sales is $4,285.80, which is 2.36 percent. Price is therefore worth 1.61 times volume and 2.64 times variable cost, and the gap between the best and the worst of the four levers is $7,018.08 of profit on the same one percent of effort.

The following figure, Figure 1.1, draws the same four levers as bars, in dollars of operating profit gained, with each lever's share of the baseline operating profit printed above its bar. The bars are ordered from the strongest lever to the weakest, and the point of the drawing is the distance between the first bar and the last one, which is the price of pulling the wrong lever for a year.

Figure 1.1 — The four profit levers of Hydrosentinelle inc. compared on year 3. Four bars measure what a one percent improvement in each lever is worth in dollars of operating profit, against a baseline year of $1,130,388 of revenue, a 62.09 percent gross margin, $520,000 of operating costs and $181,808 of operating profit. A one percent price improvement is worth $11,303.88, or 6.22 percent of the operating profit; one percent more volume is worth $7,018.08, or 3.86 percent; one percent off fixed operating costs is worth $5,200.00, or 2.86 percent; one percent off variable cost is worth $4,285.80, or 2.36 percent. Price is 1.61 times volume and 2.64 times variable cost. Hydrosentinelle inc. is fictional.

The following table, Table 1.3, gives the same four comparisons in figures and adds the column the bars cannot carry: the reason each lever is or is not the one an owner reaches for. Its columns give the lever, the one percent move, the dollars of operating profit gained, the share of the baseline operating profit, and the practical reason the lever is pulled or left alone. The last column is where the chapter's argument actually lives, because the ranking of the first four columns is arithmetic and the behavior in the fifth is not.

Table 1.3 — The four profit levers on year 3 of the case

1.3.3 What the lever does and does not promise

The arithmetic above holds volume constant, and that assumption has to be stated out loud, because a price increase that loses customers is not worth one percent of revenue. The honest way to use the lever is as a measure of what is at stake, not as a prediction. The question it answers is "how much is a point of price worth to us," and the answer for Hydrosentinelle is $11,303.88 a year. The question it does not answer is "how much volume can we afford to lose," and that one has a precise arithmetic answer too, which Chapter 2 computes as the break-even volume change of a price move: moving the subscription from $79 to $89 would remain profitable even if it cost 13.51 percent of subscription volume. That is the number that turns an argument between two cofounders into a decision.

KEY CONCEPT Price is the only decision that improves profit without requiring the business to do anything else Every other profit improvement obliges the business to change an activity. More volume must be produced, delivered and supported, and it must usually be bought with marketing or selling effort first. A lower variable cost must be negotiated with a supplier who has interests of its own, or engineered out of a process at the cost of time. A higher price requires no new activity at all: the same units, the same people, the same suppliers, the same delivery, and more money. That is why a point of price is worth more than a point of anything else, and it is also why the decision feels so exposed. The business is simply asking for more, and the purpose of this book is to make sure the asking is justified before it is made.

1.4 The Challenges: Cost Anchoring, Fear and the Absence of a Policy

1.4.1 Cost anchoring: letting the supplier set the price

The most common pricing failure in small business is not a bad number; it is a bad starting point. An owner who begins with cost and adds a percentage has built a machine that converts supplier decisions into customer prices, and the machine runs in both directions. When a supplier raises a price, the business raises its own and tells the customer that costs went up, which is a true statement and a weak argument.

Cost anchoring also produces a characteristic blindness. Because the method is internally consistent and always produces an answer, it never signals that it is wrong; a business can apply it for a decade and never learn that its buyers would have paid half again as much, because the method never asks. Chapter 2 shows, on this case, that a value-based reading of what a building manager avoids puts the ceiling far above the price Hydrosentinelle charges: the company captures 33.66 percent of the economic value it creates, and it arrived at that figure by accident.

WATCH OUT A price set from cost is a price set by your supplier A cost-plus price delegates the most important decision in the business to parties who are not in the business. The supplier sets the input cost, the payroll market sets the labor cost, and the owner sets only a percentage that was probably learned from somebody else's trade. Three consequences follow, and all three are common. First, an efficiency gain becomes a customer discount: the business works out how to install a kit in four hours instead of six and immediately gives the saving away. Second, the price becomes impossible to defend, because "our costs went up" tells the buyer about the seller's problems and nothing about the buyer's. Third, the business can never discover that it is underpriced, because the method has no input that could reveal it. Cost belongs in pricing in exactly one place: as the floor below which the business declines the work.

1.4.2 Fear, and what it is actually made of

Ask an owner why the price has not moved and the answer is usually some version of "we'd lose customers." Examined, that fear is almost always three separate fears wearing one coat. The first is fear of the conversation: nobody enjoys telling a long-standing customer that the price is going up, and the discomfort of the conversation gets mistaken for the risk of the decision. The second is fear of the unknown quantity: the owner genuinely does not know how many customers would leave, and in the absence of a number the imagination supplies a catastrophic one. The third is fear of being found out — a suspicion that the price is not really justified and that raising it will expose the fact.

Each of the three has a different remedy; the conversation is a communication problem, and Chapter 8 supplies the notice period and the sequence that makes it survivable. The unknown quantity is an arithmetic problem, and Chapter 2's break-even volume change turns it into a number: for the case, 13.51 percent of subscription volume could be lost before a move from $79 to $89 stopped paying, and Chapter 8 estimates the actual loss at 6 subscriptions beyond normal churn against a tolerable 28. The suspicion of being unjustified is an evidence problem, and Chapter 2's value analysis is the evidence. Fear decomposed into three named problems with three named remedies is no longer a reason to do nothing.

1.4.3 The absence of a policy, which is the real cause

Behind cost anchoring and fear sits a structural absence. Most small businesses have no pricing policy, by which this book means a written document of one or two pages stating the list prices, the margin floor below which the business does not sell, who may approve what discount, what a discount must be traded for, and how often prices are reviewed. Without that document, every one of the thousand small pricing decisions a business makes in a year is made alone, under pressure, by whoever is in the room, with no record.

The absence is expensive in a way that never appears on any statement. A discount given on a Tuesday afternoon to close a deal before a weekend does not show up as a line called "margin given away"; it shows up as a slightly lower gross margin rate, spread across the year, attributable to nothing. Chapter 6 makes it visible by building the pocket price waterfall of a single deal, and Chapter 9 makes it visible across a whole year by reading the distribution of discounts already granted.

1.4.4 The two symptoms that bring most readers to this book

Book E1 (The Sales Process: From Prospect to Signed Contract), Chapter 14, sends readers to this book when it observes two symptoms, and they are the two most common reasons anyone opens a book about pricing. The first is discounting everywhere: every deal closes below list, the exceptions have become the rule, and no one can say what the real price of the business is. The second is options that nobody chooses: the business presents three choices and buyers take the same one every time, which means the other two are doing no work and one of them is probably making the chosen one look expensive.

Neither symptom is a pricing method problem, which is why both are named here rather than buried. Discounting everywhere is a policy problem: there is no floor and no authority, so Chapter 6 supplies both and Chapter 9 shows how to read the discounts already granted. Options that nobody chooses is a structure problem: the tiers are not separated by anything a buyer can perceive, so Chapter 5 shows how to build a set where each option is chosen by somebody, and Chapter 3 explains why a third option changes what buyers pick. A reader who arrived with one of these two symptoms should start at Chapter 6 or Chapter 5 and come back to Chapter 2 afterwards.

1.5 What a Business Needs from Its Pricing

1.5.1 Six needs, not one decision

"Fixing the pricing" sounds like one task and is six. A business needs, first, a floor it will not sell below, expressed as a margin rather than a discount percentage, so that the worst deal it does is still a deal it wanted. It needs, second, a reason for its price that can be said out loud in two sentences and that refers to the buyer's situation rather than the seller's costs. It needs, third, a structure — tiers, units, options, minimums — that lines up with the ways its buyers actually differ, so that different buyers pay different amounts for legitimate reasons. It needs, fourth, a policy: who may discount, by how much, and in exchange for what. It needs, fifth, a calendar, because a price that no scheduled event ever touches will drift downward in real terms forever. And it needs, sixth, a measurement that says whether any of this worked, which means at minimum the win rate, the average discount and the margin actually realized.

The following table, Table 1.4, lists the six needs with the question each one answers, the evidence each one requires before it can be written down, and the chapter that supplies it. Its columns run from the need to the chapter, and a reader can use it as a checklist by marking which of the six currently exists in writing.

Table 1.4 — The six things a business needs from its pricing

1.5.2 Starting from what you already have

A business that has none of the six is not starting from nothing, because it already holds the data set that matters most and has never looked at it. That data set is its own history of quotes and sales, which records what was asked, what was given away and what was won, and therefore contains the business's real price distribution. A business of this size cannot test its way to a price; it can read its own past with discipline, and that is what Chapter 9 asks it to do.

METHOD Reading your own pricing in one afternoon When to use it: before choosing any pricing method, and again at the start of every annual price review, so that the work begins from what the business actually does rather than from what it believes it does. Steps: (1) Pull the last twelve months of closed sales into one sheet, one row per transaction, with the list price, the invoiced price and the date. (2) Add the unit cost of each transaction, taken from the accounting records rather than from memory, and compute each row's margin in dollars and as a percentage of price. (3) Sort by margin rate and read the bottom decile out loud; these are the deals the business would have been better off declining, and there will be more of them than expected. (4) Compute the difference between total list value and total invoiced value, divide by total list value, and write the result as the discount actually granted over the year. (5) List every concession that never appeared on an invoice — freight absorbed, extra visits, unbilled overruns, extended payment terms, free items — and estimate each in dollars. (6) Count the quotes sent and the quotes won and compute the win rate, noting it beside the average discount. (7) Write one page containing the numbers from steps 4, 5 and 6 and the bottom-decile list from step 3, and give it to the person who negotiates. Deliverable: a one-page reading of the business's real pricing over twelve months, with the average discount granted, the off-invoice concessions in dollars, the win rate and the deals below the intended margin. Verification criterion: the invoiced total on the page reconciles to the revenue in the accounting records for the same period, within rounding, and every concession estimated in step 5 is traceable to at least one named transaction.

1.5.3 The margin floor is the one to write first

Of the six needs, one has priority. A floor is the only one that protects the business while the other five are still being built, and the only one that can be written in an afternoon with figures the business already has. Hydrosentinelle's floor, set in Chapter 6, is $3,500 for a kit, which is 27.08 percent below the list price of $4,800 and still leaves $1,400 of margin on a $2,100 cost. Writing that single number changes every negotiation the company conducts, because the question in the room stops being "how low can we go" and becomes "what do we get in exchange for going lower."

WHO DOES THIS Who owns the price in a solo business, a small team and an established company In a solo business, the owner owns the price and also gives it away, which is the hardest configuration to manage because there is no second person to hold the line. The remedy is to separate the two roles in time rather than in people: set the price and the floor on a scheduled day, in writing, and treat that written floor as binding during negotiations, exactly as an employee would. In a small team, the price should be owned by the person accountable for the margin and never by the person whose income depends on closing the deal; in Hydrosentinelle's case the price belongs to Laurence Bergeron, the president, while Karim Haddad, who leads sales and operations, holds an authority limit and a set of trades rather than a free hand. Chapter 11 develops these roles, and Chapter 6 sets the authority limits that make them operational.

1.5.4 What the business must be able to hold, before it buys anything

A pricing capability is a set of functions before it is a set of tools, and the functions can be enumerated without naming a single product. The business must be able to hold a catalog in which every sellable thing has a price and a cost recorded beside each other, produce a document that presents a price with its scope, options, taxes, terms and expiry date, bill a recurring charge without human intervention, record every discount granted with its author and its reason, compute a margin by transaction and by segment, keep the policy where the people who negotiate can find it, date every price change, and export all of the above.

TOOLS OF THE TRADE What a pricing stack must do, before any tool is chosen A pricing stack is judged on eight criteria, not on features. Catalog with cost: every sellable item carries a price and a cost side by side, in the same record. Price list: the current prices are readable in one place by everyone who quotes, with an effective date. Quotation: a document can be produced with scope, options, taxes, payment terms, a deposit and an expiry date. Billing: recurring charges are issued without human intervention. Discount capture: every reduction from list is recorded with its amount, its author and its justification, including the reductions that never appear on an invoice. Margin analytics: results can be sliced by transaction, product, segment, channel and seller. Governance: the policy, the floor and the authority limits sit where the people who negotiate will actually look, and every price change is dated and attributable. Export: everything above can be extracted in an open format. Chapter 12 maps the nine functions of a full stack against what a spreadsheet, a market tool and La Machine each cover.

IN LA MACHINE Where a price lives in La Machine today In La Machine, a price is a field on three kinds of object, and each of the three carries a cost field beside it. A session package (Business → Session Packages, business/packages and productivity/packages) holds a name, a type of either meetings or minutes, a quantity, an unlimited option, a price, a cost and a currency, and the page computes "Profit per sale" and "% margin" on screen without storing either result. A subscription package (business/subscription-packages) holds a name, a price, a cost, a description, an image, an active flag and a billing period that may be monthly or annual and nothing else. A store product (Business → Store → Products, business/shop and business/shop-product-edit) holds a price, a currency, a status, a product cost, a shipping cost and shipping fees, and displays a profit panel computed from them. Revenues & Expenses (Account → Profile → Revenues & Expenses, business/subscriptions) is the only view in the platform that places a sale's revenue beside its cost lines. What is not here: there is no price list, no price history and no effective date, and /dashboard shows demonstration widgets with no revenue, margin or discount indicator of any kind.

1.6 The Constraints: Law, Competition, Capacity, Cost and Data

1.6.1 The law constrains what you may say about a price

The first constraint on a price is not economic. In Canada, the federal Competition Act governs representations made to the public, and a price displayed on a page, printed in a quote or spoken to a buyer is a representation. The Act has two parallel regimes. Part VII.1 treats a false or misleading representation as reviewable conduct, which is civil: no intention has to be proved, there is no criminal record, and the remedies include orders and administrative monetary penalties. Section 52 treats the same conduct as a criminal offense when the representation is made knowingly or recklessly.

Two provisions bear directly on how a price may be displayed. Subsection 74.01(1.1) makes it reviewable conduct to display a price that is not attainable because of fixed obligatory charges added later, with an exception for amounts imposed by an Act of Parliament or of a provincial legislature — sales taxes, in practice. Subsection 52(1.3) is the criminal mirror of the same rule. Section 74.01 also governs comparisons with an ordinary price, which Chapter 6 treats in detail because that is where promotions go wrong. The practical rule that Book B1, Chapter 6, states for the whole library is the one this book keeps: the price the buyer sees first is the price the buyer pays, apart from taxes, and every crossed-out price is real and documented.

THE RULE Competition Act, s. 52: a knowingly false or misleading price representation is a criminal offense in Canada Jurisdiction: Canada, federal. Instrument: Competition Act, RSC 1985, c C-34, s. 52, in the consolidation current to July 21, 2026 and last amended March 26, 2026, read on the Justice Laws Website on September 19, 2026. Subsection 52(1) provides that no person shall, for the purpose of promoting a product or a business interest, knowingly or recklessly make a representation to the public that is false or misleading in a material respect; subsection 52(1.2) adds that making a representation includes permitting one to be made. Subsection 52(1.3) provides that the display of a price that is not attainable because of fixed obligatory charges is false or misleading, except where those charges represent only an amount imposed by or under an Act of Parliament or of a provincial legislature. Subsection 52(4) provides that the general impression conveyed, as well as the literal meaning, is taken into account, which means a statement can be technically accurate and still unlawful. Subsection 52(5) provides that a person convicted on indictment is liable to a fine at the discretion of the court and to imprisonment for up to fourteen years, and on summary conviction to a fine of up to $200,000 and imprisonment for up to one year. Scope for a reader of this book: every price a business publishes, quotes, advertises or lets a system display on its behalf. Verify the current text before relying on it.

WHAT CHANGED The 2022 amendments made an unattainable displayed price an express matter Before 2022, a business that advertised one price and collected a larger one because of mandatory fees added at the end had to be pursued under the general provisions on false or misleading representations, which required the regulator to establish that the representation was misleading in the circumstances. The 2022 amendments to the Competition Act added subsections 52(1.3) and 74.01(1.1), which state directly that a price is false or misleading when it is not attainable because of fixed obligatory charges, with an exception only for amounts imposed by law. The Competition Bureau explains on its page on the practice, updated August 24, 2026, that since those provisions were added it no longer has to demonstrate the deceptive character of the display itself, and that a business should never advertise a price a consumer cannot actually obtain (Competition Bureau Canada, 2026). The practical consequence, as of September 19, 2026, is that a mandatory administration or service fee revealed only at the payment step is an express problem rather than an arguable one.

1.6.2 Québec adds a display rule of its own

A business selling to consumers in Québec is subject to a second regime, and it is stricter about display than the federal one. The text of sections 223 and 224 of the Consumer Protection Act could not be read at source on September 19, 2026, because LégisQuébec refused access; section 224 is cited here through the reading made for Book E1 on September 8, 2026, at the consolidation of April 7, 2026, and this book puts no words between quotation marks that it has not read.

QUÉBEC & CANADA NOTE The advertised price is the total the customer must pay, apart from taxes In Québec, the Consumer Protection Act, CQLR c P-40.1, section 224, provides that a merchant may not, in an advertisement, give less prominence to the total price of an assembly of goods or services than to the price of a part of that assembly, may not disclose the amount of installments without disclosing and giving prominence to the total price, and may not charge a higher price than the advertised price. That is the reading made at source for Book E1 on September 8, 2026, at the consolidation of April 7, 2026; LégisQuébec returned an access error on September 19, 2026, and the text was not read again for this book. The Office de la protection du consommateur, the body that enforces the Act, states on its page of February 10, 2025 that the advertised price must correspond to the total amount the customer has to pay: administration fees, transport and preparation fees, and charges arising from legal obligations that are not remitted to a public authority must all be included, while the GST, the QST and specific contributions remitted to a public authority may be added at payment; the same page adds that the total must be displayed more prominently than its components (Office de la protection du consommateur, 2025). Chapter 7 shows how the rule applies to a quotation with GST and QST stated separately.

1.6.3 Competition constrains the reference, not the price

The second constraint is the set of alternatives a buyer can see. A competitor's price does not determine what a business may charge, but it does determine what the buyer compares against, and a business that ignores the comparison is not avoiding it; it is only failing to participate in it. Chapter 2 builds that comparison on the case and shows that the nearest competitor's annual cost of $1,368 is not the same purchase as Hydrosentinelle's $1,908. A competitor's price also constrains in a legal direction that surprises owners: discussing prices with a competitor, however informally, engages section 45 of the Competition Act, which treats an agreement between competitors to fix, maintain, increase or control prices as a criminal offense with no need to prove any effect on competition. Chapters 6 and 11 treat that boundary.

1.6.4 Capacity constrains in both directions

The third constraint is the business's own capacity to deliver. A business that is fully booked and still quoting at the same price is not being competitive; it is rationing by queue rather than by price, and a queue is the least profitable rationing device available. Book E1, Chapter 10, says the same thing from the sales side: when capacity is the binding constraint, the remedies lie in delivery or in the offer, and raising the price is one of the legitimate remedies in the offer. This book therefore treats a capacity constraint as one of the sound reasons to raise a price, alongside cost inflation and a genuine improvement in the offer, and Chapter 8 develops all of them.

1.6.5 Cost constrains the floor, and nothing above it

The fourth constraint is cost, and it has exactly one legitimate role: it establishes the point below which the business declines. Book G1 treats the computation of cost and margin as a matter of accounting and this book uses the result. What cost may not do is generate the price.

1.6.6 Data constrains what a small business can know

The fifth constraint is the least discussed and the most consequential for readers of this book. Large firms set prices with experiments because they have enough transactions for an experiment to produce an answer, and a business with a hundred quotes a year does not. Chapter 9 computes the sample Hydrosentinelle would need for an honest test of a 5 percent effect on its win rate and finds 1,477 quotes per arm, 2,954 in total, or about 30.8 years of quoting. This is not a reason to abandon evidence; it is a reason to use the kinds that work at this volume: the business's own transaction history, the distribution of discounts already granted, field tests on real prices to real buyers, conjoint analysis on a modest sample, and structured analysis of the deals lost.

The following table, Table 1.5, gathers the five constraints, what each one actually limits, what each one is often wrongly believed to limit, and the chapter that treats it. Its columns run from the constraint to the chapter, and the third column is the one to read carefully, because most pricing paralysis comes from believing that a constraint forbids more than it does.

Table 1.5 — The five constraints on a price

1.7 What Pricing Is Not Responsible For

1.7.1 A loss to a structurally cheaper model is a targeting signal

The most useful thing a first chapter can do is prevent a reader from misdiagnosing the problem. Book E2 (Prospecting and Outbound Sales), Chapter 2, makes the point precisely: a deal lost on price to a competitor whose business model is fundamentally cheaper tells the business about a segment it should exclude, not about its pricing. This book agrees and supplies the test. A loss is a pricing signal when the winning competitor offers substantially the same outcome, at substantially the same cost to serve, at a lower price; the business is then leaving margin or losing deals it should win. A loss is a targeting signal when the winning competitor reaches the outcome by a structurally different route — less service, no installation, self-configuration, offshore labor, a subsidized adjacent revenue stream — so that its cost base is different in kind.

The practical version of the test is a question: if we had won this deal at the competitor's price, would we have wanted it? When the answer is yes, the business has a pricing problem and Chapter 2 or Chapter 6 will help; when it is no, the business has a targeting problem and Book E2, Chapter 2, is the right reading.

1.7.2 Pricing cannot fix an offer nobody wants

A price is an amount asked for a thing, and if the thing is not wanted, no amount makes it wanted. Lowering the price of an offer that fails to solve a problem produces cheaper failure and destroys the ability to raise the price later, because the low price becomes the reference. The design of the offer itself is the subject of Book A6 (Designing a Value-Added Offer); this book takes the offer as given and puts a price on it.

1.7.3 Pricing is not the plan, the pitch or the invoice

Three neighboring activities are frequently confused with pricing, and each belongs to another book. Writing the offer and pricing section of a business plan is the subject of Book A2 (The Business Plan), Chapter 5: that chapter presents a price in a document intended for a reader who is evaluating the business, which is a different exercise from deciding what the price should be. This book decides; A2 presents, and neither reopens the other's work. Presenting a price to a buyer in a conversation, defending it against an objection and trading a concession for a commitment is the subject of Book E1, Chapters 7 and 8, and that boundary is the narrowest in the library: A7 builds the options and the floor, and E1 presents and defends them. Chapter 3 of this book explains the mechanism behind the anchor and the three-option set that E1 uses, Chapter 5 explains how those options are constructed and what margin each carries, and Chapter 7 covers the document and the price page; none of the three covers the conversation. Finally, issuing the invoice and collecting the money is the subject of Book G6, while this book sets the payment terms, because a deposit and a net period are pricing decisions.

1.7.4 Pricing is not accounting, cash management, tax or contract drafting

Four more boundaries complete the map. Computing the cost of goods, reading the statements and presenting the margin is the work of Book G1, and this book consumes its output. Managing the gap between paying suppliers and being paid by customers is the work of Book G2; this book decides whether to ask for a deposit and what an annual prepayment discount really costs, and G2 manages the resulting cash. Administering the GST and the QST is the work of Book G5, while this book says only how taxes appear in a displayed price and in a quotation. And drafting the price, indexation and renegotiation clauses of a contract is the work of Book H1 (Contracts and Commercial Agreements), Chapters 3, 7 and 9; this book decides whether to index and computes what indexation is worth, and H1 writes the clause that makes the decision enforceable.

1.8 How This Book Is Organized

1.8.1 Four decisions, then time, measurement and practice

Most owners believe pricing is one decision. It is four, and separating them is the organizing idea of this book. The level is how much: Chapter 2 supplies the three classic methods and the arithmetic that chooses among them, and Chapter 3 supplies the psychology that governs how a level is perceived. The structure is how the price is shaped: Chapter 4 chooses the revenue model — one-time, recurring, usage-based, freemium, licensing, commission or outcome-based — and Chapter 5 builds the tiers, units, minimums and options within it. The policy is what may be given away and by whom: Chapter 6 supplies the floor, the authority limits, the rule of exchange and the law of promotions. The communication is how the price reaches the buyer: Chapter 7 covers the price page, the quotation, the taxes, the deposit and the expiry date.

Two chapters then handle time and evidence. Chapter 8 treats change over time: why a static price falls, what justifies an increase, how to size it, how much loss is tolerable, who keeps the old price and how indexation works. Chapter 9 treats measurement: the six numbers a pricing dashboard needs, price realization, win rate, margin by segment, the distribution of discounts, and the ways of measuring willingness to pay that do not require asking the question. Chapters 10 to 13 close the book in the standard order of this encyclopedia: the tasks and the cadence, the roles and the outside help, the tools, and a thirty-day plan.

The following table, Table 1.6, groups the thirteen chapters by the decision each settles and states the question a reader typically arrives with. Its columns give the chapters, the decision they settle and the reader's question, and it is meant to be used for navigation rather than read straight through.

Table 1.6 — The thirteen chapters, grouped by the decision each settles

1.8.2 The running case

Hydrosentinelle inc. is a fictional company in Sherbrooke that designs and sells connected water-leak detectors for owners and managers of commercial and rental buildings. A customer buys an installed kit of sensors and a gateway and pays a monthly subscription for monitoring and alerts. The company was founded by Laurence Bergeron, the president, who leads product and engineering, and Karim Haddad, who leads sales and operations; Book A5 (Financing a Business) followed the same company through its financing, and this book picks it up in its third year with the prices A5 recorded unchanged: a kit at $4,800 against a cost of $2,100, a subscription at $79 a month against a cost of $15, an annual prepayment at a 10 percent discount, and 206 buildings under subscription at the start of the year.

Over the course of this book the company does what the book teaches. It measures the economic value it creates for a building manager and finds that it captures 33.66 percent of it. It replaces a single price with three tiers, lifting the average revenue per building from $79 to $112. It builds the pocket price waterfall of one negotiated deal and finds that 30.59 percent of its list margin left in six lines, none of which was decided as a discount. It writes a floor of $3,500 and an authority limit of 6 percent for a representative and 12 percent for the head of sales. And it raises the subscription from $79 to $89 with sixty days of notice and six months of grandfathering, for a net gain of $19,392 in a full year. Every figure comes from one model, which imports its starting point from the model of Book A5 and verifies it by assertion, so that the two books cannot disagree.

1.8.3 Reading paths

A solo owner who suspects the price is too low should read Chapters 1, 2 and 8, then Chapter 13. An owner who discounts on every deal should read Chapters 1, 6 and 9 in that order, and come back to Chapter 2 once the floor exists. A team building or rebuilding an offer structure should read Chapters 4, 5 and 3 together, because the revenue model, the tiers and the perception of the tiers are one design problem. An owner facing a cost increase should read Chapters 8 and 2. A reader who arrived from Book E1, Chapter 14, with options that nobody chooses should start at Chapter 5 and then read Chapter 3. Everyone should read Chapter 6, because the policy is the part of pricing that protects all the rest of it.

Costly Mistakes

• Treating the price as a property of the product. A price is a decision made by a person on evidence, and a business that speaks of "our price" as a fact has quietly stopped deciding it.
• Building the price from the cost. A cost-plus price hands the most important decision in the business to suppliers, gives away every efficiency gain to customers who never asked for it, and can never reveal that the business is underpriced.
• Confusing the four levers and pulling the weakest. Cutting variable cost by one percent is worth $4,285.80 on the case while a point of price is worth $11,303.88, and the effort required runs in the opposite direction.
• Running a business with no written floor. When two people in the same business give different answers to "what is the lowest price we accept," the business does not have a floor; it has two opinions and a spread of prices it cannot explain.
• Diagnosing a targeting problem as a pricing problem. A loss to a competitor with a structurally cheaper model is a signal to exclude a segment, and matching that competitor's price moves the loss from the sales report to the income statement.
• Advertising a price that the buyer cannot actually obtain. A mandatory fee added at the payment step is an express matter under the Competition Act since the 2022 amendments, and sales tax is the only charge that may lawfully be added to a displayed price.

Key Takeaways

• A price is the amount asked for a defined offer, on defined terms, from a defined buyer, at a defined moment, and it is a decision rather than a property of the thing sold.
• Cost, value, revenue, margin and pocket price are five different things, and almost every pricing error comes from asking one of them a question only another can answer.
• Cost sets the floor, competition sets the reference and value sets the ceiling, and reversing that order is the most common structural mistake in small-business pricing.
• Price is the strongest of the four profit levers in any business whose gross margin is below one hundred percent; on the case, one percent of price is worth $11,303.88, against $7,018.08 for volume, $5,200.00 for fixed costs and $4,285.80 for variable cost.
• Two-thirds of the smallest Canadian businesses expected their prices to stand still while nearly two-thirds expected a cost obstacle, which is a forecast of margin compression rather than of stability.
• A business needs six things from its pricing — a floor, a reason, a structure, a policy, a calendar and a measurement — and the floor is the one to write first.
• Five constraints bound a price — law, competition, capacity, cost and data — and the law constrains what may be said about a price rather than how high it may be.
• Pricing is not responsible for an unwanted offer, a targeting mistake, a sales conversation or a collection problem, and treating any of those as a pricing problem makes all of them worse.

Your Action Plan

1. Write down the four numbers that rank your levers. The deliverable is one line containing your last full year of revenue, variable cost, gross margin and operating costs; it takes about thirty minutes with your accounting records, and it is done when the four numbers reconcile to your income statement for the same period.
2. Compute your four levers. The deliverable is a four-row table like Table 1.3, showing what one percent on each lever is worth in dollars and as a share of your operating profit; it takes about fifteen minutes once step 1 exists, and it is done when the four amounts are ranked and the ranking is written in a sentence.
3. Name how your current price was actually set. The deliverable is one paragraph identifying which of the six routes in Table 1.2 produced your main price, and what evidence, if any, supported it; it takes about twenty minutes, and it is done when a colleague who was present at the time agrees with your account.
4. Read your own pricing over twelve months. The deliverable is the one-page reading described in the Method callout, containing the average discount granted, the off-invoice concessions in dollars, the win rate and the deals below your intended margin; it takes one afternoon, and it is done when the invoiced total on the page reconciles to your recorded revenue.
5. Write a provisional margin floor for your main offer. The deliverable is a single number, expressed as a margin in dollars or as a margin rate, with the unit cost it is based on; it takes about an hour, and it is done when everyone who quotes has been told the number and can repeat it from memory.
6. Audit your displayed prices against the law. The deliverable is a list of every place your price appears publicly, with a note beside each saying whether the first number a buyer sees is the total payable apart from taxes; it takes about two hours, and it is done when every mandatory charge you add is either included in the displayed price or imposed on you by statute.
7. Put the pricing work on a calendar. The deliverable is a recurring annual review date and a quarterly margin read, each with a named owner and a definition of done; it takes about thirty minutes, and it is done when both appear in the system the business actually uses, with an owner and a due date.

In La Machine

Every route named in this section was verified in the route inventory on September 19, 2026 and read in the product's source code at commit be8bf04 of September 7, 2026 (La Machine, 2026); nothing below is inferred from a page title.

• Business → Session Packages (business/packages, also reachable at productivity/packages) holds a session package with a name, a type of either meetings or minutes, a quantity, an unlimited option, a price, a cost and a currency, and the page computes "Profit per sale" and "% margin" on screen from the price and the cost without storing either value.
• business/subscription-packages holds a subscription package with a name, a price, a cost, a description, an image, an active flag and a billing period that may be monthly or annual and nothing else.
• Business → Store → Products (business/shop and business/shop-product-edit) holds a store product with a price, a currency, a status, a product cost, a shipping cost and shipping fees, and shows a profit panel computed from them; the comparison price is displayed but is not saved.
• Account → Profile → Revenues & Expenses (business/subscriptions) reads subscription transactions, package purchases and product purchases, and writes each sale as a revenue line with its cost lines beneath it; it is the only view in the platform that places a revenue beside its margin.
• Control center → Plan tasks and Control center → Today hold the recurring pricing work of step 7 of the action plan as scheduled tasks with an owner and a due date, and Control center → Audit log records who changed what.
• What is not here. La Machine holds no price list and no price catalog, no price history and no effective date on a price change, no margin floor, no discount approval and no discount record, no quotation of any kind — the page at business/quotes is a page of motivational quotes — no invoice for the reader's own customers, no price test, no elasticity and no competitor. The /dashboard shows demonstration widgets with no revenue, margin or discount indicator, and productivity/tools offers an arithmetic calculator, a stopwatch, timers and a timesheet, with no financial calculator. The platform checks no price presentation against the Competition Act or the Consumer Protection Act; the rule that the price seen first is the price paid is the reader's to apply.

Self-Check

1. What are the five words that are not the price, and what does each one describe?
2. Why is a one percent price improvement worth more than a one percent volume improvement in almost every business, and what does the comparison assume?
3. What did the four levers come to on year 3 of the case, in dollars and as a share of operating profit?
4. What are the six things a business needs from its pricing, and which one should be written first?
5. How do you tell a pricing problem from a targeting problem when a deal is lost on price?

Answers. (1) Cost is what the business gives up to produce and deliver the offer; value is what the offer is worth to the buyer against the buyer's best alternative; revenue is price multiplied by quantity; margin is price minus cost, in dollars or as a percentage of price; and pocket price is what is actually kept after every discount, concession, extended term and unbilled overrun (Sections 1.1.2 and the Definition callout). (2) Because a price improvement adds its full amount to profit with no additional cost incurred, so it is worth one percent of revenue, while a volume improvement must be produced and delivered and is therefore worth only one percent of gross margin; the comparison assumes that volume does not fall in response to the price, and Chapter 2 supplies the break-even volume change that tests the assumption (Sections 1.3.1 and 1.3.3). (3) Price was worth $11,303.88, or 6.22 percent; volume $7,018.08, or 3.86 percent; fixed costs $5,200.00, or 2.86 percent; and variable cost $4,285.80, or 2.36 percent, against a baseline operating profit of $181,808 (Section 1.3.2, Table 1.3 and Figure 1.1). (4) A margin floor, a reason for the price, a structure, a discount policy, a calendar and a measurement; the floor should be written first, because it protects the business while the other five are still being built (Sections 1.5.1 and 1.5.3). (5) By asking whether the winning competitor reaches the same outcome with a similar cost to serve, in which case it is a pricing problem, or by a structurally different route with a different cost base, in which case it is a targeting signal and the segment should be excluded (Section 1.7.1).

Chapter References

Competition Bureau Canada. (2026, August 24). Drip pricing. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/en/deceptive-marketing-practices/drip-pricing

Competition Act, RSC 1985, c C-34, s. 52 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Competition Act, RSC 1985, c C-34, s. 74.01 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Consumer Protection Act, CQLR c P-40.1, s. 224 (Québec). LégisQuébec, consolidation of April 7, 2026, as read for Book E1 on September 8, 2026; not read again, LégisQuébec returning an access error on September 19, 2026.

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Marn, M. V., & Rosiello, R. L. (1992, September–October). Managing price, gaining profit. Harvard Business Review. https://hbr.org/1992/09/managing-price-gaining-profit (article page and publisher product description read September 19, 2026; the full text was not read, and the pagination reported by secondary sources could not be verified at source)

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Office de la protection du consommateur. (2025, February 10). Annonce d'un prix tout inclus. Gouvernement du Québec. Retrieved September 19, 2026, from https://www.opc.gouv.qc.ca/commercant/pratique-commerce/publicite-loi/prix-tout-inclus

Statistics Canada. (2026a, June 18). Analysis on small businesses in Canada, second quarter of 2026 (Catalogue no. 11-621-M). Retrieved September 19, 2026, from https://www150.statcan.gc.ca/n1/pub/11-621-m/11-621-m2026008-eng.htm

Chapter 2 — Pricing Foundations

Hydrosentinelle enters year 3 with two prices that nobody has re-examined since the first year: a kit at $4,800 and a monitoring subscription at $79 a month. Both were set the same way, and the way is the most common one in small business. The founders added up what a kit cost them, $2,100 of hardware and installation, roughly doubled it, rounded the result to a number that sounded deliberate, and went back to building the product. Chapter 1 measured what that habit is worth in year 3: one percent on price is $11,303.88 of operating profit, against $7,018.08 for one percent on volume, on a total operating profit of $181,808. This chapter asks the question Chapter 1 left open. Where should the number come from in the first place? There are three honest answers — the cost of making the thing, the price of the alternatives the buyer can see, and the value the thing creates for the buyer — and each answers a different question. The chapter takes the three apart and gives the one calculation that makes any price move decidable: the change in volume at which the move breaks even. That calculation replaces a question nobody can answer, how much demand will I lose, with one anyone can answer: can I afford to lose that much?

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Say which question each of the three classic methods answers, so that you can use cost-plus, competition-based and value-based pricing together instead of choosing one and defending it.
• Compute a cost-plus price correctly, and never again confuse a markup on cost with a margin on price.
• Build a defensible comparison with competitors by putting every offer on the same annual basis, instead of comparing one visible number with another.
• Estimate the economic value your offer creates for a specific customer, name the reference price it is measured against, and state what share of that value your price captures.
• Compute the break-even volume change of any price move in one line, and read it as the loss you can tolerate or the volume you must win.
• Choose a method for a given offer with a written rule, and know what evidence you need before you can use the strongest of the three.

Why It Matters

Most owners believe they have one pricing decision to make, and that the decision is a number. They have at least four, and the number is the last of them. The first is the method: what kind of fact the price will rest on. The second is the level, which is where in the band the number sits. The third is the structure, which Chapter 5 treats, and the fourth is the policy that governs what happens to the number in a negotiation, which Chapter 6 treats. The method comes first because it decides which facts the business must go and collect: a business that prices from cost needs a good cost model and nothing else, while a business that prices from value needs to know what happens in its customer's operation when the product works, which is much harder to learn and far more valuable to own.

For a solo business, the method decides whether the owner's income grows with skill or stays pinned to an hourly cost. For a small team, it stops price from being an argument between the person who fears losing the sale and the person who signs the checks. For an established business, the method is usually implicit and inherited, and the cost of leaving it implicit compounds: prices set from cost in an expensive year stay high after materials fall, and prices set by matching a competitor stay matched long after that competitor stopped being the relevant alternative. Book B1 (Marketing Fundamentals for Business Owners), Chapter 6, named the four methods this library would treat — cost-plus, competition-based, value-based and subscription — and promised that the methods behind the numbers would live here. Three are in this chapter; the subscription is a revenue model as much as a price, and Chapter 4 treats it.

There is a second reason this chapter comes early. A price formed from cost is a price set by your suppliers, and a price formed from fear is a price set by the most anxious person in the room. Neither is a decision. Book A1 (From Idea to Business Model), Chapter 7, told the founder to form a first price hypothesis from evidence rather than from cost or from fear, and to test it without asking people what they would pay. This chapter takes that hypothesis further: it gives the arithmetic that turns it into a defensible number, and the calculation that says whether moving it is worth the risk.

In This Chapter

• 2.1 The Three Classic Methods and the Question Each One Answers
• 2.2 Cost-Plus: Mechanism, Arithmetic and Failure Modes
• 2.3 Competition-Based: Reference Prices and Their Traps
• 2.4 Value-Based Pricing and Economic Value to the Customer
• 2.5 Price Elasticity of Demand
• 2.6 The Break-Even Volume Change of a Price Move
• 2.7 Willingness to Pay: What It Is and Where It Lives
• 2.8 Choosing a Method: A Decision Rule

2.1 The Three Classic Methods and the Question Each One Answers

2.1.1 Three methods, three different questions

The three classic methods of setting a price are not three ways of producing the same number. They are answers to three different questions, and the confusion between them is the source of most bad pricing in small business.

Cost-plus pricing answers the question, what must I charge not to lose money? It starts from the cost of producing and delivering one unit and adds a markup meant to cover fixed costs and leave a profit. Its input is internal and knowable, which is why almost every business starts there.

Competition-based pricing answers the question, what will the buyer compare my price with? It starts from the prices the buyer can see — a competitor's published price, a substitute, an in-house solution, doing nothing at all — and positions the business's price relative to them. Its merit is that it takes the buyer's point of view seriously: a buyer never sees a price in isolation, but always beside something else.

Value-based pricing answers the question, what is this worth to the buyer? It starts from what changes in the buyer's operation, measured in the buyer's own money, when the offer works, and takes a share of that change. Its input is hard to collect, which is why it is the least used and the most profitable of the three. The pricing literature is unambiguous that value should lead, costs setting a floor rather than a price (Nagle et al., 2023), and Book A2 (The Business Plan), Chapter 5, already told the reader that the answer to "why this price and not less" is never "our costs."

KEY CONCEPT Cost sets the floor, competition sets the reference, value sets the ceiling Every price sits inside a band with three walls, and each of the three classic methods measures one of them. The floor is the cost of serving one more customer; below it, every sale makes the business poorer, and no volume repairs it. The ceiling is the economic value the offer creates for the buyer compared with the buyer's best alternative; above it, a well-informed buyer is better off doing something else. The reference is the price the buyer will compare yours with, which is usually the most visible alternative rather than the best one. The three walls are measured separately and none can be inferred from another: a high cost does not create value, a low competitor's price does not lower your cost, and a large value does not mean the buyer knows about it. The pricing decision is the choice of a point inside that band, and the argument the business makes for that point. When the floor is above the ceiling, no price works, and the answer is not a clever number but a different cost structure, a different segment, or no offer at all.

2.1.2 What each method gets right, and how each one fails

Each method has a characteristic success and a characteristic failure. Cost-plus gets the floor right and is defensible internally; it fails because it carries no information about the buyer, and because it is circular for any business with fixed costs, the unit cost depending on a volume that depends on the price. Competition-based pricing gets the reference right; it fails when the business copies a competitor whose costs and strategy differ, and hardest of all when the observed competitor is not the buyer's real alternative. Book E2 (Prospecting and Outbound Sales), Chapter 2, makes that point from the sales side: a deal lost on price to a competitor with a fundamentally cheaper model is a signal about the segment you chose, not about your price. Value-based pricing gets the ceiling right and produces the strongest argument a seller can make; it fails when value is asserted rather than measured, when the buyer does not believe it, and when value varies so much between customers that one price cannot serve them, which is a structure problem for Chapter 5 rather than a level problem.

The table that follows compares the three methods on five points: the question each one answers, the input it needs, where it typically fails, and the section that develops it. Read it as a map of the chapter, since the three are used together and in the order the last column implies.

Table 2.1 — The three classic methods, and the question each one answers

2.2 Cost-Plus: Mechanism, Arithmetic and Failure Modes

2.2.1 The mechanism, and the two numbers it requires

Cost-plus pricing has two inputs and one operation. The inputs are the unit cost and the markup; the operation is multiplication. At Hydrosentinelle, the unit cost of a kit is $2,100, made of $1,650 of hardware and $450 of installation labor, as Book A5 (Financing a Business), Chapters 1 and 2, established when the company was financed. A markup of 100 percent on that cost gives a price of $4,200. The company charges $4,800, which is a markup of 128.57 percent on cost, because the founders rounded up to a number that looked deliberate. The rounding was worth more than any pricing analysis the company did that year: $4,800 − $4,200 = $600 per kit, and at the 180 kits the model plans for year 3, $600 × 180 = $108,000, which is 59.40 percent of the year's entire operating profit of $181,808. That is not a recommendation to round upward. It is a measure of how little the cost-plus number had to do with the value the company was delivering, and of how much money sits between the floor and the ceiling when nobody measures the ceiling.

The subscription shows the same thing more sharply. The monthly service costs $15 to deliver, so the same rule of doubling the cost would have priced it at $30. The company charges $79, a markup of 426.67 percent, and the model's margin rate on that line is 81.01 percent. Nobody at Hydrosentinelle could have defended $79 from the cost. The founders arrived at it because a competing system's monitoring fee was in that range, which is competition-based pricing applied without anyone naming it.

The first input hides most of the difficulty. A unit cost is not a fact a business discovers but a construction that depends on decisions: whether the installer's travel sits in the kit cost or in overhead, and whether the warranty claims that arrive eighteen months later are counted at all. A cost model that omits them produces a floor that is too low, and a business that prices from a floor that is too low loses money on exactly the sales it is proudest of winning.

2.2.2 A markup is not a margin, and the difference is expensive

The most common arithmetic error in small-business pricing is the confusion between a markup on cost and a margin on price. They are two ratios with the same numerator and different denominators: a markup divides the dollars added by the cost, a margin divides the same dollars by the price. A 50 percent markup is a 33.33 percent margin, and a 100 percent markup is a 50 percent margin. An owner who intends a 40 percent margin and applies a 40 percent markup earns 28.57 percent, on every unit, for as long as the error survives.

The table that follows makes the conversion concrete on the kit's cost of $2,100. Its four columns give the markup applied to the cost, the price that results, the margin in dollars and the margin as a percentage of the price. Read the last two rows together: the company's actual price of $4,800 corresponds to a markup of 128.57 percent and a margin of 56.25 percent, and the gap between those two percentages is the whole point of the table.

Table 2.2 — The same kit cost of $2,100, priced at four markups

WATCH OUT A markup is not a margin Two ratios are built from the same dollars, and mixing them costs money quietly. A markup is the amount added divided by the cost; a margin is the same amount divided by the price. On Hydrosentinelle's kit, $2,700 added to a $2,100 cost is a 128.57 percent markup and a 56.25 percent margin. The error has a predictable shape: an owner who wants a 40 percent margin and multiplies the cost by 1.40 gets 28.57 percent, losing more than a quarter of the intended margin on every unit. It runs the other way in negotiation, where a salesperson told never to go below a 30 percent margin, applying 30 percent to the cost, is authorizing 23.08 percent. Three habits prevent it. State every internal target as a margin on price, because that is the number that appears in the statements Book G1 (Accounting Fundamentals for Business Owners) explains. Write both ratios side by side in the margin model. And when a supplier quotes a standard markup for your industry, ask which denominator they mean before you use it.

2.2.3 Which costs belong in the floor

The floor that matters for a single pricing decision is the incremental cost: everything the business would not spend if it did not make this sale, and nothing else. For the kit, that is the $1,650 of hardware plus the $450 of installation. The president's salary, the office, the software subscriptions and the certification work already paid for are not in it, because they do not change with one more kit.

The distinction matters when the business is deciding whether to accept an unusual order at a reduced price, because any price above the incremental cost adds something toward the fixed costs. It is dangerous when the exception becomes the rule: a price above the incremental cost but below the full cost makes the business poorer as soon as those sales become normal. The safe formulation is that the incremental cost is the floor for an exceptional decision, and full cost recovery at planned volume is the floor for the price list. Chapter 6 builds the case's negotiated floor on that reasoning and places it at $3,500, well above the $2,100 incremental cost and well below the $4,800 list price.

The circularity of cost-plus appears here, because allocating fixed costs per unit requires a volume, and the volume depends on the price being set. The practical resolution is to allocate on a planned volume that the business commits to, to state that volume in writing beside the price, and to revisit the price when the actual volume differs materially. Hydrosentinelle's year-3 plan carries 180 kits, an average of 281 subscriptions and $520,000 of fixed operating costs, and those three numbers are what the company's allocations rest on.

WHO DOES THIS Who computes the cost, and why the accountant must agree The cost model is the one piece of pricing work that must not be done alone by the person who sets the price. In a solo business, the owner builds it and an accountant reviews it once a year, because an owner's cost model almost always omits the owner's own time and the costs that arrive late, such as warranty claims and rework. In a small team, the operations lead owns the variable cost of delivery, the person who buys owns the input costs, and the bookkeeper or accountant reconciles the model against the financial statements at least quarterly; that reconciliation is the control, because a margin model that cannot be tied to the income statement is a spreadsheet of opinions. In an established business, a controller owns the cost model and the pricing owner consumes it, and the two roles are deliberately separate. The accountant's agreement matters for one reason above all: the margin rate defended in a negotiation must be the margin rate that appears in the statements, or the business will discover at the year-end that it defended a floor that was never a floor.

2.2.4 The failure modes, and where cost-plus is still the right answer

Cost-plus fails in four recognizable ways. It imports your supplier's decisions, since your price rises when hardware costs rise whether or not your buyer's value has changed. It exports your efficiency gains, since your price falls when your process improves. It ignores the buyer entirely, producing the same price for a customer who would have paid double and for one who will never buy. And it anchors the whole organization on cost, so that every conversation about price becomes a conversation about cost, including the conversations with customers.

Cost-plus is nonetheless the right method in three situations. It is right for commodity work where the buyer genuinely cannot distinguish between suppliers. It is right where a contract or a regulation requires it, as in some public procurement. And it is right as a floor check on every other method, which is how this book uses it: no price produced by value or by competition is accepted until its margin has been compared with the floor.

WHAT CHANGED Since June 23, 2023, an agreement between employers about pay is a criminal offense in Canada A cost-plus price rests on a labor cost, and Canadian law changed what an employer may do about that cost with other employers. Subsection 45(1.1) of the Competition Act, added by the 2022 amendments and in force on June 23, 2023, makes it a criminal offense for unaffiliated employers, whether or not they compete for customers, to agree to fix, maintain, decrease or control wages, salaries or terms and conditions of employment, or to agree not to solicit or hire one another's employees. The Competition Bureau's enforcement guidelines state that it applies to agreements entered into on or after that date and to later conduct that reaffirms or implements an earlier one; that "terms and conditions of employment" is read broadly, covering benefits, hours, location and non-compete clauses; that only reciprocal no-poaching agreements are caught; and that the ancillary restraints defense of subsection 45(4) is recognized for mergers, joint ventures, franchises and service contracts, while clauses clearly broader than necessary may still be investigated (Competition Bureau Canada, 2023). The penalty is imprisonment for up to 14 years and a fine at the discretion of the court. For an owner building a cost model, the consequence is narrow: benchmark labor costs from published wage data and your own hiring experience, never from an understanding with another employer, and treat a trade-association conversation about pay scales the way you would treat one about prices. Verify the current text before relying on it.

2.3 Competition-Based: Reference Prices and Their Traps

2.3.1 What a reference price is, and why the buyer always has one

A buyer never evaluates a price in the abstract. Every price is read against a reference: another supplier's quotation, a remembered price for something similar, the cost of the in-house workaround, or the cost of doing nothing. Competition-based pricing is the deliberate choice of which reference the business will be read against, so it is not imitation but the management of a comparison.

The first question is therefore not what competitors charge but what this buyer will compare you with. For Hydrosentinelle, a property manager considering leak detection compares the offer with a competing connected system, with a plumbing inspection contract that has no sensors, and with doing nothing and carrying the risk; only the first is a competitor in the ordinary sense, and in many lost deals the winner was the third.

The second question is what to compare. Hydrosentinelle sells a kit plus a monthly subscription, so a buyer who compares list prices is comparing two numbers that mean different things over different periods. The only honest common basis is the annual cost of ownership, the purchase amortized over the equipment's useful life, which the model sets at five years, plus twelve months of subscription. On that basis, Hydrosentinelle costs $4,800 ÷ 5 + $79 × 12 = $960 + $948 = $1,908 a year.

2.3.2 The three competitors of the case, on one basis

The model carries three competitors. They are anonymized and fictional, invented for the case and named Competitor A, Competitor B and Competitor C; no real supplier is described, compared or ranked anywhere in this book, and their figures are assumptions chosen to make the arithmetic of the comparison visible.

The table that follows puts the four offers on the same annual basis. Its five columns give the offer, the one-time kit price, the monthly fee, the annual cost of ownership computed as the kit amortized over five years plus twelve monthly fees, and the reduction in water-damage losses each supplier claims. The point of the table is in the fourth column: the rankings you would form from the first two columns do not survive it.

Table 2.3 — Four offers on one basis: annual cost of ownership for one building

Two readings follow, and each changes a decision. First, which number the buyer looks at decides who looks expensive. On the kit alone, Hydrosentinelle is 23.08 percent more expensive than Competitor A and 22.58 percent cheaper than Competitor C; on the monthly fee alone, it is 61.22 percent above Competitor A and 38.76 percent below Competitor C; on the annual cost of ownership, it is 39.47 percent above Competitor A and 17.33 percent below Competitor B. A buyer who fixates on the kit price and a buyer who fixates on the monthly fee are having two different conversations with the same seller, and Chapter 7 shows how a quotation should present both.

Second, the cheapest offer is not the reference for everyone. Competitor A is the natural reference for a buyer who has decided that all these systems are the same; for a buyer who has read the loss-reduction claims, the comparison set is Hydrosentinelle and Competitor C, which claim the same 60 percent, and on that basis Hydrosentinelle is 31.56 percent cheaper. A competitor's price is in any case data about that competitor, not about you: it may reflect lower costs, a weaker product, a decision to buy market share, or a mistake.

CASE IN POINT The three competitors Hydrosentinelle is compared with Fictional example; names and figures are invented. Karim Haddad, who leads sales and operations, kept a one-page sheet of the three systems the company met in deals. The sheet had four lines per competitor: the published kit price, the published monthly fee, the annual cost of ownership on the five-year basis, and the loss-reduction claim the competitor made in its own marketing. Every figure came from a public source — a published price page, a brochure collected at a property-management trade show, or a quotation a prospect chose to share after the decision — and carried the date it was read. Competitor A sold a $3,900 kit with a $49 fee, $1,368 a year, claiming a 35 percent reduction in losses; Competitor B a $5,600 kit with a $99 fee, $2,308 a year, claiming 55 percent; Competitor C a $6,200 kit with a $129 fee, $2,788 a year, claiming 60 percent. Hydrosentinelle's own $1,908 sat between A and B, with the same 60 percent claim as the most expensive of the three. The cofounder's conclusion changed the sales conversation for the rest of the year: the company had been losing deals to Competitor A on the kit price while holding the strongest position in the market on annual cost per point of loss reduction, and nobody had said so out loud. Two rules governed the sheet. Every figure came from a public source or from a customer, never from a conversation with a competitor. And the sheet recorded what buyers would compare, not prices to match.

2.3.3 The traps of competition-based pricing

Five traps recur. The first is comparing incomparable numbers, which Table 2.3 corrects by moving everything to an annual basis. The second is treating a list price as a transaction price, since Chapter 6 shows a real transaction price falling 17.21 percent below list in a single deal. The third is following the wrong competitor, usually the loudest rather than the one your buyers consider. The fourth is matching a price without matching the offer, which turns a comparison into a promise the business has not costed. The fifth, the most expensive, is using the competitor's price as the ceiling, which holds only if the two offers are identical to the buyer, and Section 2.4 is the remedy.

There is a sixth trap, and it is not a business mistake but a criminal offense. The easiest way to know a competitor's price is to ask the competitor, and that conversation is the one thing a business may never have.

THE RULE Competition Act, s. 45: an agreement with a competitor about price is a criminal offense in Canada Under section 45 of the Competition Act, R.S.C. 1985, c. C-34, it is a criminal offense to conspire, agree or arrange with a competitor or potential competitor, in respect of a product, to fix, maintain, increase or control its price; to allocate sales, territories, customers or markets; or to fix, control, prevent, lessen or eliminate its production or supply. The prohibition is per se: no effect on competition needs to be proven, and subsection 45(3) allows the agreement to be inferred from circumstantial evidence, the criminal standard of proof still applying. Subsection 45(8) defines "price" as including discounts, rebates, allowances and price concessions, so an understanding about discount levels is an understanding about price. Subsection 45(4) provides an ancillary restraints defense where a restraint is directly related to, and reasonably necessary for, a broader lawful agreement between the same parties, and subsection 45(6) excludes agreements between affiliated persons. The penalty on indictment is imprisonment for up to 14 years and a fine at the discretion of the court, with no statutory ceiling since the 2022 amendments. Scope and date: federal, applying throughout Canada to businesses of every size, read on the Justice Laws Website at the consolidation current to July 21, 2026, last amended March 26, 2026. In practice: gather competitor prices from published pages, brochures, public tenders and customers who volunteer them, never from a competitor, a shared spreadsheet or a trade-association discussion; leave a meeting where prices come up, in a way others notice; and call a lawyer the same day if you believe you may have been part of such a discussion. Verify the current text before relying on it.

2.3.4 Gathering competitor prices honestly

A competitor price file is worth keeping, and it is easy to keep lawfully. Four sources are legitimate and sufficient: published prices, captured with the date they were read; public procurement records, which publish awarded amounts in many jurisdictions; buyers, who often volunteer what they were quoted, bearing in mind that a prospect's account of a rival quotation is a negotiating move as often as a fact; and public statements in annual reports and the trade press. Two sources are closed: a conversation with a competitor about prices is the offense described above, and obtaining confidential pricing through someone who owes that competitor a duty of confidentiality creates exposure of a different kind, which Book H1 (Contracts and Commercial Agreements) treats. The file stays small — published prices, date read, source, scope, outcome claims — and Chapter 10 puts its twice-yearly review on the calendar.

QUÉBEC & CANADA NOTE Comparing a Québec price with a competitor's across the border A Canadian owner who compares prices with a supplier selling from the United States must normalize three things before the comparison means anything. Currency comes first: a price in US dollars converts at a rate that moves, so the file records the amount, the currency and the conversion date, and a quotation honored for thirty days carries either a currency clause or a margin for the movement, which Chapter 7 treats. Tax and duty come second: an imported system may carry customs duty and brokerage on top of the sales taxes, so the landed cost rather than the invoice price is what the buyer compares, and Book G5 (Business Taxes in Québec and Canada) treats the taxes themselves. The display basis comes third, and this is where Québec differs. The Office de la protection du consommateur states that a price advertised to a consumer must correspond to the total amount the customer must pay, that administration fees and transport and preparation charges must be included, that only the GST, the QST and specific contributions remitted to a public authority may be added at the cash, and that the total must be more prominent than any of its components (Office de la protection du consommateur, 2025). The Office administers the Consumer Protection Act; its pages express how it applies the Act rather than the text of the Act itself. Two consequences follow. Hydrosentinelle sells to property managers, so its quotations are business-to-business and the consumer display rules do not govern them, while the federal Competition Act does. And the moment the same company publishes a price page a consumer can buy from, the Québec requirement applies to that page in full. Verify the current text before relying on it.

2.4 Value-Based Pricing and Economic Value to the Customer

2.4.1 The idea, stated exactly

Value-based pricing sets the price from what the offer is worth to the buyer rather than from what it costs the seller or from what a rival charges. The sentence is easy to say and easy to get wrong, because "worth" has to be made precise before it becomes arithmetic: worth compared with what, to whom, over what period?

The precise version is this. For a specific buyer, compared with that buyer's best available alternative, the offer is worth the money the buyer saves plus the money the buyer gains by using it instead of that alternative, over a stated period. That sum is the economic value to the customer, usually shortened to EVC, and it has two parts: the price of the reference alternative, which the buyer would have paid anyway, and the differentiation value, which is everything the offer does that the alternative does not, converted into the buyer's money and net of anything the offer does worse.

Three consequences follow. Because EVC is measured against an alternative, a change in the alternative changes the value of your offer without anything in your offer changing. Because it is measured for a specific buyer, there is no such thing as the value of your offer in general. And because it is measured in the buyer's money, the estimate is built from the buyer's operating facts, not from the seller's belief about how good the product is.

DEFINITION Economic value to the customer, reference value and differentiation value The economic value to the customer (EVC) is the total monetary worth of an offer to a specific customer over a stated period, measured against that customer's best available alternative. It has two parts. The reference value is what the customer would pay for that best alternative, whether a competitor's product, an in-house solution, or the status quo, whose reference value is zero. The differentiation value is the net monetary effect of every difference between your offer and that alternative: costs avoided, revenue gained, time no longer spent and risk no longer carried, minus the costs your offer adds, such as switching, training and disruption. EVC is a ceiling, not a price: a rational, fully informed customer will not pay more than EVC, and will usually pay considerably less, because the customer must be left a reason to change suppliers. The share of EVC that the price takes is the value captured, and the rest is the customer's surplus. Two warnings belong to the definition. An EVC quoted without naming its reference alternative means nothing. And an EVC the customer does not believe is worth nothing in a negotiation, which is why the estimate is built from the customer's own figures wherever possible.

2.4.2 Building the estimate on the case

The model builds the estimate for one representative building, and every input is an assumption of the case, stated so that a reader can substitute their own. The building has 48 units. A water-damage event in such a building costs the owner $24,000 on average, all in, and happens once every two years, which is an expected annual loss of 0.5 × $24,000 = $12,000.

Hydrosentinelle's system reduces those losses by 60 percent, because its sensors report a leak within minutes; Competitor A, the reference, reduces them by 35 percent, reporting once a day. The difference in avoided losses is $12,000 × (60% − 35%) = $3,000 a year. The second component is insurance: on an annual premium of $16,000, the insurer grants 5 percent for a continuously monitored system and 2.5 percent for a daily-reporting one, a difference of $16,000 × (5% − 2.5%) = $400. The third is staff time: the building manager spends 20 hours a year less on inspections and small incidents, valued at the manager's loaded internal cost of $45 an hour, which is 20 × $45 = $900.

The three components give a differentiation value of $3,000 + $400 + $900 = $4,300 a year. Added to the reference value of $1,368, the annual cost of owning Competitor A's system, the economic value to the customer is $5,668 a year.

Figure 2.1 shows that construction and what the price does with it: the left column builds the $5,668 from the reference value and the differentiation value, and the right column splits the same $5,668 into the part Hydrosentinelle charges and the part that stays with the customer.

Figure 2.1 — Economic value to the customer, and the share the price captures. The left column builds the economic value of the Hydrosentinelle offer for one 48-unit rental building: it starts at the reference value of $1,368 a year, the annual cost of owning the reference competitor's system, and adds $4,300 of differentiation value to reach $5,668. The right column splits that same $5,668 between the $1,908 a year that Hydrosentinelle charges, which is 33.66 percent of the value, and the $3,760 that stays with the customer, which is 66.34 percent. The three boxes below the columns give the arithmetic of the differentiation value: $3,000 of avoided losses, $400 of insurance credit and $900 of staff time. Hydrosentinelle inc. is fictional, its competitors are anonymized and invented, and every figure is computed by the case model.

The table that follows gives the same construction line by line, for a reader who needs to rebuild it with different assumptions. Its three columns give each line of the estimate, its annual amount, and where that amount comes from. The last three lines are the ones to remember: the ceiling, the price, and the share of the ceiling the price takes.

Table 2.4 — The economic value estimate for one 48-unit building, line by line

2.4.3 Total value created is not the ceiling

A distinction that costs businesses a great deal of money hides in the first line of that table. Measured against doing nothing at all, the offer is worth $12,000 × 60% + $16,000 × 5% + 20 × $45, which is $7,200 + $800 + $900 = $8,900 a year; measured against Competitor A, it is worth $5,668. The price of $1,908 is therefore 21.44 percent of the total value created and 33.66 percent of the differential value.

Which of the two is the ceiling depends on what the buyer can actually do instead. A buyer who has never heard of Competitor A faces a ceiling of $8,900; a buyer with Competitor A's quotation on the desk faces $5,668, because $1,368 of the value can be bought elsewhere. That is why the discovery work of Book A1 (From Idea to Business Model), Chapter 7, and the qualification work of Book E1 (The Sales Process: From Prospect to Signed Contract), Chapter 7, feed the price directly.

2.4.4 What share of the value to capture

No rule of thumb about the right share survives contact with a real market, but three principles decide it. The customer must keep enough of the value to have a reason to act, because changing supplier costs time, attention and risk; the 66.34 percent left with the customer is not generosity but the budget for the buyer's decision. The share you can capture rises with the credibility of the estimate, since an estimate built from the building's own premium and loss history is worth several times one built from a brochure. And the share must survive the segment being wrong: a six-unit building with new plumbing has a much smaller EVC, and a price set at the 48-unit building's ceiling would be unsellable there. The remedy is not a lower price for everyone but a structure whose metric follows the value, which is what Chapter 5 builds.

METHOD Building an economic value estimate in six steps When to use it: before setting or defending the price of any offer whose benefit a customer could, in principle, put in dollars, which includes most business-to-business offers and many consumer ones. Steps: (1) Name one specific customer or one narrowly defined customer type, and write down the operating facts you know about it, such as size, volume, current spending and current losses. (2) Name the best alternative that customer actually has, which may be a competitor, an in-house workaround or doing nothing, and compute its cost over a stated period on the same basis as your own offer, amortizing any one-time amount over the useful life. (3) List every difference between your offer and that alternative, positive and negative, in plain sentences before any number is attached. (4) Convert each difference into the customer's money, using the customer's own figures wherever possible and writing out the arithmetic, then subtract the costs your offer adds. (5) Add the reference value and the net differentiation value to obtain the EVC, and state the period it covers. (6) Choose the share you will capture, write the resulting price, and compute the margin it earns against the floor of Section 2.2. Deliverable: a one-page estimate naming the customer type, the reference alternative, each difference with its arithmetic, the EVC, the price and the margin. Verification criterion: a customer of that type, shown the page, can say which figures come from their own operation and can dispute a number without disputing the approach.

2.4.5 From the estimate to the argument

The estimate is also the sales argument, which is the practical reason to build it even in a year when the price does not change. A defensible price comes with three sentences: what the buyer's current situation costs, what changes when the offer works, and what share of the difference the price takes. For the case: this building loses about $12,000 a year to water damage; our system removes about 60 percent of that and earns a larger insurance credit, which is about $5,668 a year of value against the system you are pricing us against; we charge $1,908, so you keep about two-thirds of the gain.

Book E1 (The Sales Process: From Prospect to Signed Contract), Chapter 7, owns the conversation in which those sentences are delivered and the presentation of options; this chapter owns what goes into them. A seller who improvises the value argument in the room will improvise a discount in the room as well, while a seller who arrives with the estimate on paper has something to trade other than the price, which is the subject of Chapter 6.

2.5 Price Elasticity of Demand

2.5.1 What elasticity measures

The price elasticity of demand measures how much the quantity sold responds to a change in price. It is the percentage change in quantity divided by the percentage change in price, and it is normally negative; this book states the sign each time, since economists report it as negative while many practitioners quote its absolute value. Demand is elastic when the absolute value is greater than one and inelastic when it is less.

The boundary at one is the boundary of revenue: at an elasticity of exactly −1, a price increase loses in volume exactly as much revenue as it gains in price. That boundary is famous and, for an owner, actively misleading, because the objective is margin rather than revenue. The units lost to an increase were carrying variable costs that are no longer incurred, so the move that leaves margin unchanged is always less demanding than the one that leaves revenue unchanged, and Section 2.6 gives the threshold that matters.

2.5.2 What makes demand elastic

Elasticity is not a property of a product. It is a property of a product in a market, for a segment, at a moment, over a range of prices. Six factors move it, and an owner can assess all six without measuring anything.

• The number and visibility of substitutes raise elasticity, because a buyer who can see three comparable offers can act on a price difference immediately.
• The share of the buyer's budget that the purchase represents raises elasticity, because buyers spend attention in proportion to the amount at stake.
• Switching costs lower elasticity, since a buyer who would have to migrate data or retrain staff will absorb an increase rather than pay those costs again.
• Urgency and the cost of failure lower elasticity, because a buyer facing an imminent loss is buying insurance rather than shopping.
• Who pays matters, since a purchase reimbursed by an insurer or an employer responds far less to price than one paid from the buyer's own pocket.
• Price visibility and the ease of comparison raise elasticity, which is why the metric chosen in Chapter 5 and the presentation decisions of Chapter 7 change the elasticity a business faces without changing the product.

2.5.3 Why a small business cannot look up its elasticity

Published elasticity estimates exist for many categories, and this book prints none of them, for two reasons.

The first is that a category average does not transfer to one business. Published estimates are averages across firms, markets, periods and price ranges, and the dispersion around them is wide; a number that describes an industry says nothing reliable about one supplier's differentiated offer to one segment at one price point.

The second is that measuring your own elasticity requires variation you do not have. Chapter 9 computes what it would take here: detecting a five-point change in the quote win rate would require 1,477 quotations in each branch of the test, about 2,954 in all, roughly 30.8 years of quoting at this volume. A small business that believes it has run such a test has usually measured the season or a single large customer.

The conclusion is not that elasticity is useless but that it must be used the other way around: the owner decides what volume change the business can tolerate, then asks whether a response that large is plausible.

TOOLS OF THE TRADE What a margin model must compute Before any pricing software, a business needs one spreadsheet that it owns and understands, and that spreadsheet must compute nine things. It must hold one unit cost per offer, separating variable from fixed and reconciling to the financial statements. It must show both the markup on cost and the margin on price, so that nobody converts a ratio in their head. It must give the contribution margin in dollars and as a rate, because that rate drives every other calculation here. It must compute the break-even volume change of any price move, as a percentage and as a number of units. It must hold the floor for each offer and flag any price below it, which is what Chapter 6 governs. It must put competitor offers on a common annual basis, with the date each figure was read and its source. It must hold the economic value estimate for each customer type, with its reference alternative named. It must compute the annual rate implied by a payment discount, because an annual prepayment discount is financing, as Book A5 (Financing a Business), Chapter 2, establishes and Chapter 4 develops. And it must be auditable: one input cell per assumption, no number typed twice, and a date on every external figure. A spreadsheet does all nine, which is why Chapter 12 treats it as a legitimate permanent tool; what it does badly is govern who may change a price, and that is a different function.

2.6 The Break-Even Volume Change of a Price Move

2.6.1 The question, reversed

Every price decision meets the same wall. Raising the price earns more on each sale and probably loses some sales; lowering it wins some sales and earns less on each. Deciding by forecasting the response would need the elasticity a small business cannot measure, so the question is reversed: how much would volume have to change for this move to leave me exactly as well off as before? That number is computable exactly from two inputs the business already has, and it converts an unanswerable forecast into an answerable judgment.

2.6.2 The formula, and where it comes from

Let the contribution margin ratio be m, the margin on price after variable costs, expressed as a decimal, and let d be the price change as a decimal fraction of the current price, positive for an increase and negative for a discount. The break-even change in unit volume is:

break-even volume change = −d ÷ (m + d)

The derivation takes two lines. Before the move, each unit contributes p × m. After the change, the price is p × (1 + d) and the variable cost is unchanged, so each unit contributes p × (m + d). For total contribution to be unchanged, Q₂ × p × (m + d) = Q₁ × p × m, so Q₂ ÷ Q₁ = m ÷ (m + d), and the change is that ratio minus one, which is −d ÷ (m + d).

Two properties deserve to be stated in words. The lower your margin, the more volume a discount must produce: a business at a 20 percent margin that discounts 10 percent must double its volume, while the same discount at an 80 percent margin needs 14.29 percent more. And the lower your margin, the more volume loss a price increase can survive, because the increase adds proportionally more to a thin unit margin. An owner who knows only one thing about their own numbers should know their contribution margin rate.

2.6.3 The four moves of the case

The model computes the break-even volume change for four moves that this book actually makes. The kit carries a contribution margin of 56.25 percent and the subscription 81.01 percent, and those two rates give very different answers to the same percentage change.

The table that follows gives the four moves. Its five columns give the move, the price change as a percentage, the contribution margin of the offer concerned, the break-even volume change from the formula, and the same figure expressed in units at the volumes the year-3 plan carries. A negative volume figure is a loss the business can absorb; a positive one is additional volume the business must find.

Table 2.5 — The break-even volume change of four price moves

Each row rewards a slow reading. A ten percent discount on the kit, the concession a salesperson grants to close a hesitant buyer before lunch, requires 21.62 percent more kits merely to stand still: at $4,320 the unit margin falls from $2,700 to $2,220, and holding the year's kit contribution of 180 × $2,700 = $486,000 takes 218.9 kits. Nobody believes that a discount enlarges the pipeline by a fifth, so it is a decision to earn less, and it should be made as such, in exchange for something, under the rule Chapter 6 writes. A five percent increase, by contrast, can lose 8.16 percent of volume, about fifteen kits, before the company is worse off.

The third row connects this chapter to Chapter 8, where the subscription rises from $79 to $89, an increase of 12.66 percent that breaks even as long as no more than 13.51 percent of the 206 subscriptions are lost, which is about 28. That chapter runs the increase with sixty days of notice and six months of grandfathering and loses six. The company did not need to know its elasticity; it needed to know that twenty-eight was the number that mattered and that six was plausible. The fourth row surprises owners of high-margin services: a ten percent discount on the subscription needs only 14.08 percent more subscriptions, less than the kit's 21.62 percent, precisely because the margin is higher, so high margins make discounts cheap in volume and expensive in dollars.

DO THE MATH The break-even volume change of a ten percent discount Fictional example; the figures are Hydrosentinelle's, invented and computed by the model. A property manager asks for 10 percent off the $4,800 kit, and the salesperson wants to know what that costs before answering. Step one, the contribution margin. The kit costs $2,100 to deliver, so the margin is $4,800 − $2,100 = $2,700, which is $2,700 ÷ $4,800 = 56.25 percent. Step two, the price change. A ten percent discount is d = −0.10. Step three, the formula. The break-even volume change is −d ÷ (m + d) = 0.10 ÷ (0.5625 − 0.10) = 0.10 ÷ 0.4625 = 21.62 percent. Step four, the check in dollars. At $4,320 the unit margin is $2,220; the year's plan of 180 kits earns 180 × $2,700 = $486,000, and holding that figure takes $486,000 ÷ $2,220 = 218.9, call it 219 kits, which is 39 more kits. Step five, the reading. The discount is worth granting only if it brings in about 39 additional kits a year, or if it buys something else worth $480 on this order, such as a second building, a two-year commitment, a reference or a faster payment, which is the rule of exchange of Chapter 6. The same arithmetic on the other three moves: a 5 percent increase on the kit tolerates a volume loss of 8.16 percent, about 15 kits; the subscription's move from $79 to $89, an increase of 12.66 percent on a margin of 81.01 percent, tolerates a loss of 13.51 percent, about 28 of the 206 subscriptions, and Chapter 8 loses only 6; and a 10 percent discount on the subscription requires 14.08 percent more subscriptions. One line to remember: with a margin of m and a price change of d, volume must change by −d ÷ (m + d), and if that number is not plausible in your market, the move is not a price decision but a wish.

2.6.4 Reading the same number as an elasticity threshold

The break-even volume change can be restated as the elasticity at which the move is exactly neutral, by dividing it by the price change. That threshold is −1 ÷ (m + d), and it is the only elasticity number a small business needs, because it is judged against rather than measured. For the four moves, the ten percent kit discount breaks even at −2.16, the five percent kit increase at −1.63, the subscription's rise from $79 to $89 at −1.07, and the ten percent subscription discount at −1.41.

Each decision then becomes one question: is my demand more or less responsive than this threshold? For a monitoring subscription embedded in a building, a response as strong as −1.07 is implausible over a single ten-dollar step, which is why Chapter 8's increase was defensible before it was attempted. For a kit whose headline price stands 23.08 percent above the most visible competitor, a response as strong as −2.16 is not obviously implausible, and that is the honest reason the company stopped discounting the kit.

2.6.5 What the formula deliberately ignores

The formula answers one question exactly and says nothing about four others. It ignores capacity, so a volume increase that requires a second installation crew is not free. It ignores the competitive response, and a discount matched within a month buys no volume while lowering the market's prices. It ignores customer lifetime, so a discount that acquires a five-year subscriber may be worth more than the unit arithmetic shows, and Chapter 4 supplies that view. And it ignores which customers are lost, since losing the fifteen most price-sensitive buyers is a different event from losing the fifteen largest accounts, which Chapter 9 measures. The arithmetic fixes what a move costs; the judgment decides whether the required change is plausible.

2.7 Willingness to Pay: What It Is and Where It Lives

2.7.1 What willingness to pay is, and how it differs from value

The willingness to pay of a buyer is the highest amount that buyer would actually hand over rather than take the alternative. It is not the economic value of Section 2.4, and the gap between them is where most pricing disappointment lives. EVC is what a fully informed, perfectly rational buyer would pay; willingness to pay is what a real buyer, uncertain about your claims and constrained by a budget set last November, will actually sign. Four things push it below EVC: doubt that the promised effect will occur in this building, switching and disruption costs the buyer counts and the seller forgets, budget authority, which caps what a manager may approve alone, and inertia.

Willingness to pay is also not a number but a distribution across buyers, and a price selects from it. A single price of $4,800 sells to every buyer whose willingness to pay exceeds $4,800 and to nobody else: the buyers just below are lost, and those far above are served at a discount they never asked for. Everything in Chapter 5 about tiers, metrics and fences exists to serve more of that distribution without lowering the price for everyone.

2.7.2 Where willingness to pay actually lives

Willingness to pay lives in behavior, and four sources of behavior are available to a small business without any research budget.

• What the buyer already spends on the problem is the strongest and most neglected source, since a property manager who pays for inspections, carries a deductible and has settled two water claims has already revealed a budget.
• The discounts already granted are a natural experiment that has run. Chapter 9 reads the case's year of quotations by discount band and finds the win rate rising from 22.58 percent at no discount to 46.15 percent above ten percent, while the margin per quotation sent peaks in the 1 to 5 percent band.
• Field tests on real prices, with a different price in one region, channel or quarter and the outcome recorded, are the only instrument that measures behavior rather than opinion, and Chapter 9 sets out how small the question must be to be answerable at this volume.
• Deposits and pre-orders are the top rung of the evidence ladder, as Book A1 (From Idea to Business Model), Chapter 7, puts it: four deposits at a price are evidence about that price, and a hundred survey answers are not.

2.7.3 The instruments that ask, and what they are worth

Two research instruments ask buyers about price without asking the naive question, and both come with limits attached.

The price sensitivity meter asks four questions — at what price the offer would be so cheap that you would doubt its quality, at what price a bargain, at what price expensive but still worth considering, and at what price too expensive to consider — and reads an acceptable range from where the answer curves cross (Van Westendorp, 1976). It measures perception rather than behavior and is sensitive to the sample and the wording, so it is useful for bounding a range that a field test will then probe, and worthless as a way of choosing a price.

Conjoint analysis shows buyers whole offers that differ in several attributes, price among them, and asks them to choose between offers rather than to price one. Because the buyer trades price against features, it recovers what each feature is worth in money and is much harder to game than a direct question; it needs a designed set of choices and a sample large enough to estimate the trade-offs, which is why Chapter 9 treats it as an instrument bought once, for a decision worth the expense.

The naive question — what would you pay for this? — is worth less than nothing, because the answers are systematically wrong in both directions and arrive with a confidence that makes them hard to discard.

2.7.4 Neither cost nor fear

Two sources of a price masquerade as methods and are neither. Cost produces a number with no information about the buyer in it. Fear produces the number the most anxious person in the room can say out loud without flinching, which is why the same offer is quoted at different prices by different people in the same company, and why the quoted price falls as the quarter ends. Both are recognizable by one test: ask what evidence would change the number. A cost-based price changes when a supplier's invoice changes, a fear-based price when the person quoting has had a bad week, and a price built from this chapter when the buyer's operation, the alternatives or the measured win rate changes. Only the third is a pricing method, and Book A1, Chapter 7, said as much before this book existed.

2.8 Choosing a Method: A Decision Rule

2.8.1 The rule

The decision rule of this chapter is about sequence and evidence rather than preference. Compute the floor from the incremental cost, for every offer. Put the visible alternatives on one common basis and identify the reference the buyer will use. Estimate the EVC against that reference for each customer type. Then let the level be decided by the strongest method the evidence supports: value when the differentiation value can be measured in the buyer's own figures and communicated credibly, the reference when the buyer reads your offer and a visible alternative as substitutes, and cost only when neither is available or a contract requires it. Finally, check the number against the floor and the move against the break-even volume change of Section 2.6. The rule has a corollary that owners resist: when the evidence for value pricing does not exist, the answer is to go and get it, not to fall back on cost forever, and that evidence is a handful of conversations with customers about what the problem costs them.

The table that follows turns the rule into a grid. Its four columns give the situation the business is in, the method that should decide the level, the evidence required before that method can be used, and the first thing to do if the evidence is missing. Read the third column as the real work: the method is easy to choose and hard to earn.

Table 2.6 — Which method decides the level, and what it requires

2.8.2 The rule applied to the case

Hydrosentinelle runs the rule on both of its offers, and the two answers differ. The kit carries a headline price 23.08 percent above the most visible alternative, and buyers compare headline numbers, so the reference dominates the level; the company's work is to move the comparison onto the annual basis of Table 2.3 and to attach the value estimate that justifies the difference, the floor being intact at a 56.25 percent margin against the negotiated floor of $3,500 that Chapter 6 sets. The subscription faces switching costs, an installed base and no easy comparison, and its differentiation value is measurable, so value dominates: $1,908 a year captures 33.66 percent of an EVC of $5,668 and leaves $3,760 with the customer. That number makes the rest of the book possible, because the tiers of Chapter 5, the increase from $79 to $89 in Chapter 8 and the discount floor of Chapter 6 are all decisions about where to sit inside a band this chapter measured.

IN LA MACHINE Cost and price side by side in a package La Machine holds one piece of this chapter's arithmetic, and it is the piece most small-business software omits: the cost of an offer, recorded beside its price. A session package (Business → Session Packages, business/packages and productivity/packages) carries a name, a type of unit, a quantity, a price and a cost, and the page computes the profit per sale and the margin percentage on screen as the fields are edited, without storing either. A subscription package (business/subscription-packages) carries a name, a price, a cost, a description and a billing period that is monthly or annual. A store product (Business → Store → Products, business/shop and business/shop-product-edit) carries a price, a currency, a product cost, a shipping cost and shipping fees, with a profit panel beside them. Because the cost is there, the margin rate of Section 2.2 is visible at the moment a price is typed, and Revenues & Expenses (Account → Profile, business/subscriptions) is the one place in the platform where a sale's revenue lines and its cost lines are read together. What is not here. There is no elasticity, no break-even volume change and no price optimization; no competitor, no reference price and no place to record what a rival charges; no economic value estimate and no price list; productivity/tools offers an arithmetic calculator, a stopwatch, timers and a timesheet, and no financial calculator, so the formula of Section 2.6 is computed in a spreadsheet; and the platform checks no price presentation against the Competition Act or the Consumer Protection Act, which Book B1, Chapter 6, already stated and this book confirms. Routes verified in the route inventory on September 19, 2026, and behaviors read in the product code (La Machine, 2026).

Costly Mistakes

• Setting the price from cost and calling it a method. A cost-plus price makes your supplier the author of your number, and at Hydrosentinelle the arbitrary $600 the founders added above the doubled cost is 59.40 percent of the year's operating profit.
• Applying a markup where a margin was intended. An owner who wants a 40 percent margin and multiplies the cost by 1.40 earns 28.57 percent on every unit, quietly, for as long as the error survives.
• Comparing a competitor's kit price with your kit price. Until every offer is on one basis, such as the annual cost of ownership of Table 2.3, the comparison measures packaging rather than price.
• Estimating value without naming the alternative. Against Competitor A, the case's ceiling falls from $8,900 to $5,668, and an estimate with no reference in it is a number the buyer cannot check.
• Granting a discount without computing the volume it requires. Ten percent off the kit demands 21.62 percent more kits, about 39 a year, and nobody who grants that discount believes those kits will appear.
• Asking buyers what they would pay. Stated intentions are the weakest evidence available; behavior, deposits and the discounts you have already granted are the honest sources.

Key Takeaways

• The three classic methods answer three different questions, and a price is a choice inside the band their answers define: cost sets the floor, competition sets the reference, and value sets the ceiling.
• A markup is not a margin. The case's kit carries a 128.57 percent markup on its $2,100 cost and a 56.25 percent margin on its $4,800 price, and every internal target should be stated as a margin on price.
• Comparison is only honest on a common basis. On annual cost of ownership, the case's offer costs $1,908 a year against $1,368, $2,308 and $2,788 for the three anonymized competitors of the model.
• Economic value is always measured against a named alternative: $1,368 of reference value plus $4,300 of differentiation value gives $5,668, of which the price of $1,908 captures 33.66 percent.
• The customer's share is the budget for the customer's decision, and the 66.34 percent left with the buyer is what makes changing supplier worth the trouble.
• A small business cannot measure its own elasticity, because the experiment would take about 30.8 years of quotations at the case's volume, so the question must be reversed.
• The break-even volume change of a price move is −d ÷ (m + d): 21.62 percent more kits for a ten percent discount, 13.51 percent of subscriptions that may be lost when $79 becomes $89.
• A price formed from cost or from fear is not a decision, and the test is simple: ask what evidence would change the number.

Your Action Plan

1. Write your contribution margin rate for every offer you sell. The deliverable is a one-page list of offers with their incremental cost, their price and their margin in dollars and as a percentage of price; it takes two hours, and it is done when your accountant agrees that the costs tie to the statements.
2. Convert every internal price target into a margin on price. The deliverable is a corrected set of targets with the markup shown beside the margin; it takes thirty minutes, and it is done when nobody has to convert a ratio in their head.
3. Build the competitor sheet. The deliverable is one line per competitor with the published prices, the source, the date read and the annual cost of ownership on your basis; it takes half a day, and it is done when you can say which alternative each segment compares you with.
4. Estimate the economic value of your main offer for one customer type. The deliverable is the six-step page of Section 2.4, with the reference alternative named and each difference costed; it takes a week, including the customer conversations, and it is done when a customer of that type recognizes the figures as their own.
5. Compute the break-even volume change of the price moves you are considering. The deliverable is a small table like Table 2.5, each move expressed as a percentage and as units; it takes an hour, and it is done when each move carries a written judgment on whether that change is plausible.
6. Decide the method for each offer, in writing. The deliverable is one line per offer naming the method that decides its level and the evidence behind it, using Table 2.6; it takes an hour, and it is done when the offers with no evidence have a dated plan to obtain it.
7. Put the review on the calendar. The deliverable is a recurring task that reopens the cost model, the competitor sheet and the value estimate; it takes fifteen minutes to create, and it is done when the first occurrence appears in the calendar that Chapter 10 builds.

In La Machine

Every route in this section was verified in the route inventory on September 19, 2026, and each behavior was read in the product code at the commit of September 7, 2026 (La Machine, 2026); Book N8 will cover these pages one by one, and this book describes only what bears on a pricing method.

• Business → Session Packages (business/packages, also reachable at productivity/packages) holds a package's name, its unit type, its quantity, its price and its cost, and computes the profit per sale and the margin percentage on screen without storing them.
• Subscription packages (business/subscription-packages) hold a name, a price, a cost, a description and a billing period that is monthly or annual, which is enough to hold the two sides of the margin arithmetic of Section 2.2.
• Business → Store → Products (business/shop and business/shop-product-edit) hold a price, a currency, a product cost, a shipping cost and shipping fees, with a profit panel that reads them.
• Account → Profile → Revenues & Expenses (business/subscriptions) is the only view in the platform that reads a sale's revenue lines together with its cost lines, which makes it the closest thing to a margin report.
• What is not here. There is no elasticity, no break-even calculation and no price optimization; no competitor, no reference price and no place to record a rival's published price with its date; no economic value estimate, no price list and no price history; no financial calculator in productivity/tools, which offers an arithmetic calculator, a stopwatch, timers and a timesheet; no margin or revenue indicator on /dashboard, which shows demonstration widgets; and no check of any price presentation against the Competition Act or the Consumer Protection Act.

Self-Check

1. What question does each of the three classic methods answer, and in what order should they be used?
2. A kit costs $2,100 and sells for $4,800. What is the markup on cost, what is the margin on price, and why does the difference matter in a negotiation?
3. Why is Competitor A's annual cost of ownership of $1,368 part of the economic value estimate rather than a competing number to be beaten?
4. A ten percent discount is requested on the $4,800 kit. How much additional volume does it require, and how is that figure computed?
5. Why does this chapter refuse to print a published elasticity figure for the case, and what does it use instead?

Answers. (1) Cost-plus answers what you must charge not to lose money, competition-based answers what the buyer will compare you with, and value-based answers what the offer is worth to the buyer; value is estimated first, the reference second, and cost last, as a floor check (Sections 2.1.1 and 2.8.1). (2) The markup is $2,700 ÷ $2,100 = 128.57 percent and the margin is $2,700 ÷ $4,800 = 56.25 percent; a salesperson told to hold a 30 percent margin who applies 30 percent to the cost is authorizing 23.08 percent (Section 2.2.2). (3) Because economic value is differential: the buyer can obtain the reference system's benefits for $1,368 a year, so only the $4,300 of differentiation value is added to it, giving a ceiling of $5,668 rather than the $8,900 the offer is worth against doing nothing (Sections 2.4.1 and 2.4.3). (4) It requires 21.62 percent more kits, about 39 a year on a plan of 180, computed as −d ÷ (m + d) = 0.10 ÷ (0.5625 − 0.10) and confirmed in dollars, the unit margin falling from $2,700 to $2,220 (Sections 2.6.2 and 2.6.3). (5) Because a category average does not transfer to one business and measuring the company's own elasticity would take about 30.8 years of quotations; the chapter computes the break-even volume change and the elasticity threshold it implies instead (Sections 2.5.3 and 2.6.4).

Chapter References

Competition Act, RSC 1985, c C-34, s. 45 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Competition Bureau Canada. (2023, May 30; modified 2024, May 1). Enforcement guidelines on wage-fixing and no poaching agreements. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/en/how-we-foster-competition/education-and-outreach/enforcement-guidelines-wage-fixing-and-no-poaching-agreements

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Office de la protection du consommateur. (2025, February 10). Annonce d'un prix tout inclus. Gouvernement du Québec. Retrieved September 19, 2026, from https://www.opc.gouv.qc.ca/commercant/pratique-commerce/publicite-loi/prix-tout-inclus

Van Westendorp, P. H. (1976, September). NSS-Price Sensitivity Meter (PSM): A new approach to study consumer perception of prices [Paper presentation]. ESOMAR 29th Congress, European Society for Opinion and Marketing Research. https://ana.esomar.org/documents/nss-pricesensitivity-meter-psm- (catalogue notice read September 19, 2026; the paper itself is restricted to members of the association and was not read, so the four questions described in Section 9.6.3 are this book's own wording and not a quotation)

Chapter 3 — Pricing Psychology

The three prices that Hydrosentinelle inc. published at the start of its third year were $89, $129 and $179 a month for one building, and the argument that produced them belongs to Chapter 2 and to Chapter 5: the value the service creates for a building owner, the alternatives that owner can buy, and the cost of serving each tier. What happened next belongs to this chapter. The president, Laurence Bergeron, wanted to publish $90, $130 and $180, on the ground that a business selling engineering should not look like a discount store. The cofounder responsible for sales and operations, Karim Haddad, wanted to publish the middle price as "$4.24 a day", and wanted a fourth line on the page reading "usually $6,348, this month $5,999". Each of those three proposals is a device of pricing psychology. One is worth less than a percentage point and is erased by a loss of volume the company could not measure. One says nothing false and still exposes the company to a federal provision on partial price representations if it is displayed the wrong way. And one is an ordinary-price claim, which is the single most regulated sentence a small business can write. This chapter takes each device in turn, gives the mechanism, gives the measurement where a measurement exists, says so plainly where none does, and ends where the devices stop being technique and start being an offense.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Explain why a buyer judges a price by comparison and decide, deliberately, what your price is compared with.
• Set an anchor from the buyer's own economics rather than from your cost, knowing what the anchoring evidence does and does not support.
• Build a three-option set in which the top option is real, and compute what that option is worth to your average revenue.
• Decide whether an ending digit is worth anything to you, using the field evidence that exists and the break-even volume beside it.
• Frame a price by the day, by the unit or all-inclusive without breaking the law, by keeping the payable amount the most prominent number.
• Apply the four legal checks between a pricing device and a false or misleading representation, and keep the records that make a was-price defensible.

Why It Matters

Pricing psychology is the part of pricing that owners read about first and understand least, and it is the smallest of the four decisions this book treats. On the case, one percentage point on the price is worth $11,303.88 of operating profit in year 3, which is 6.22% of the year's EBITDA, while the whole gain from rounding three prices up to the next ten is 0.89% of the average revenue per building and disappears if volume falls by 1.11%. These devices change a few tenths of a point; the decisions in Chapters 2, 5 and 6 change tens of points. That order of magnitude is the first thing to know about the subject, and the reason this chapter refuses to sell tricks.

For a solo business, the chapter matters because the owner sets the ending digits alone, usually late at night, and usually downward; separating the level of the price from the way it is said is the only protection against that reflex. For a small team, it matters because the page, the quote and the proposal are written by different people, and because a sentence written by a seller who wants to close — "usually $6,348" — can create a legal exposure the owner never approved. For an established business, it matters because reference prices accumulate: customers remember what they paid, the market remembers what was advertised, and a business that has trained its market to wait for a promotion has changed its own reference price in a way that takes years to undo.

There is a third reason, and it is why this chapter is longer on law than most chapters on psychology. A pricing device that informs a buyer is a technique. A pricing device that misleads a buyer is a reviewable practice under the civil provisions of the Competition Act, and, if it is made knowingly or recklessly, a criminal offense under section 52 of the same Act. The line is not drawn by the seller's intention alone: the court considers the general impression conveyed as well as the literal meaning of what was said (Competition Act, s. 52(4)). Book B1 (Marketing Fundamentals for Business Owners), Chapter 6, states the practical rule this book adopts without weakening it: the price the buyer sees first is the price they pay, apart from taxes, and every was-price is real and documented. Sections 3.7 and 3.8 demonstrate that rule provision by provision.

In This Chapter

• 3.1 Why a Price Is Judged Relatively, Never Absolutely
• 3.2 Anchoring
• 3.3 Decoys and the Compromise Effect
• 3.4 Charm Prices and Price Endings
• 3.5 Framing: Per-Day, Per-Unit, Partitioned and All-Inclusive
• 3.6 Bundling and Unbundling
• 3.7 Reference Prices, Fairness and the Cost of Being Caught
• 3.8 Where Psychology Ends and the Law Begins

3.1 Why a Price Is Judged Relatively, Never Absolutely

3.1.1 The buyer has no absolute scale

A buyer who is told that a monitoring service costs $129 a month cannot say whether that number is large or small until it is placed beside something else. Nothing in the number itself carries a verdict. The verdict comes from a comparison, and the buyer will make one whether or not the seller offers one: against the price paid last year, against a competitor's quotation, against the budget line the purchase will be charged to, against the cost of the problem it prevents, or, failing anything better, against the other numbers on the same page. Every device in this chapter is a way of choosing that comparison, and every legal provision in it is a limit on how far a seller may go in choosing.

The consequence for a small business is immediate. A price published alone is not a neutral presentation; it delegates the comparison to the buyer, who will then compare it with the cheapest thing that resembles it, because that is the easiest number to find. Book E1 (The Sales Process: From Prospect to Signed Contract), Chapter 7, makes the same observation about a proposal and draws the conclusion for the sales conversation: the seller places the buyer's own cost of the problem before the price, so that the price is read as a saving rather than as an expense. That chapter presents the price; this chapter builds the comparison set the presentation uses.

DEFINITION Anchoring, decoy, compromise effect, charm price, framing, partitioned pricing and bundling Anchoring is the tendency of a number encountered first to pull later judgments of related quantities toward itself, so that the same price is judged differently depending on what was seen before it. A decoy is an option added to a set not because anyone is expected to buy it but because its presence changes how the other options are judged. The compromise effect is the related tendency of a middle option in a set of three to be chosen more often than the same option would be chosen in a set of two. A charm price is a price whose ending digits are chosen for their effect rather than for arithmetic, most often a price ending in nine; the choice of ending is called the price ending. Framing is the expression of one and the same amount in a different unit or against a different reference — by the day, by the apartment, by the seat — without changing what is owed. Partitioned pricing is the display of a price split into a base and one or more separate charges, as opposed to all-inclusive pricing, which displays a single total; when the separate charges are mandatory and are not imposed by law, a partitioned display becomes a partial price representation in Canadian law, treated in Section 3.8. Bundling is the sale of two or more items as one offer at one price, and unbundling is the separation of an offer into items priced and sold on their own. In plain terms: anchoring and decoys change the comparison, charm prices and framing change the way the number is said, and bundling changes what is being counted.

3.1.2 What the evidence supports, and what it does not

Pricing psychology attracts more confident writing than any other part of pricing, and the confidence outruns the evidence. Three things have to be kept apart: a mechanism, which explains why something might happen; a measurement, which observes how much it happened in a stated setting; and a transfer, which claims the same amount will happen elsewhere. Most pricing advice offers a mechanism, borrows a measurement from a retail setting, and asserts the transfer without argument. This chapter names the evidence it has for each device and names the gap where it has none.

The table that follows is the inventory of that evidence. Its first column names the device, the second the measurement this chapter can cite, the third what was actually measured in that source, and the fourth how far the result carries. Read the fourth column before using any device in the first: three of the six rows say, in substance, that the arithmetic here is exact and the behavioral claim behind it is not measured, which is a reason to be modest rather than a reason to do nothing.

Table 3.1 — What the evidence behind each device supports

One row of that table rests on a field experiment, and it is worth noticing how uneven even good evidence is: Anderson and Simester (2003) measured about 35% more units in one study, about 15% in another and about 7% in a third, and described one of their own comparisons as a tentative finding worth further investigation. That is a long way from the confident single number that pricing advice usually quotes. The last row carries a caution of a different kind. A meta-analysis of 1,851 price elasticities drawn from 81 studies reports an average of −2.62, meaning that a one percent price increase went with a fall in quantity of about 2.6 percent, but the corpus is dominated by consumer packaged goods measured with scanner data, and the authors also find that measured elasticities have grown in absolute size over four decades and that studies ignoring the endogeneity of price understate them (Bijmolt et al., 2005). The number is a reminder that demand moves, not a coefficient to reuse in a small business.

3.1.3 The size of the prize

Before spending an afternoon on ending digits, an owner should know what such an afternoon can be worth. On the case, the four profit levers of Chapter 1 are not close: a one-point improvement in price is worth $11,303.88 of operating profit in year 3, against $7,018.08 for a point of volume, $4,285.80 for a point of variable cost and $5,200.00 for a point of fixed cost, on a baseline EBITDA of $181,808.00. The devices of this chapter operate well below one point: rounding the ladder up to the next ten moves the average revenue per building by 0.89%, and the top tier's presence is worth 5.86% of it under an assumption about where its buyers would otherwise go. Those amounts are real and worth having; they are not a pricing strategy.

They are also small beside the amounts that leak out of a price once it has been set. Chapter 6 follows a single negotiated deal from a list price of $4,800.00 to a pocket price of $3,974.12, which is 82.79% of list, and shows that 30.59% of the list margin disappeared in six lines of which none was negotiated as a discount. Marn and Rosiello (1992) made the same point in the article that named the pocket price waterfall and the pocket price band, and the publisher's description of that article reports that a spread of more than 60% between the highest and the lowest price actually realized for the same product is common among manufacturers. A business with a spread of that size does not have a psychology problem; it has a policy problem, and Chapter 6 treats it.

KEY CONCEPT A price is a comparison, and you choose what it is compared with The idea that organizes this chapter is that a buyer never evaluates a price on its own. The buyer evaluates a difference: between this price and the last one, between this option and the next, between this cost and the cost of the problem. The seller's choice is not whether to influence that comparison, since a price page, a quote and a conversation all influence it by their structure alone, but whether to influence it deliberately or accidentally. Choosing deliberately means deciding, before the page is written, what the price should be read against: the value the buyer gains, which Chapter 2 estimates at $5,668.00 a year for the reference building against the $1,908.00 the case charges; the alternative the buyer would otherwise buy, which for the case is a competitor at $1,368.00 a year; or the cost of doing nothing. Choosing accidentally means publishing a number beside a cheaper number and letting the difference speak. Every honest device here applies that idea, and every dishonest one tries to make the comparison point at something that is not there.

3.2 Anchoring

3.2.1 The mechanism, and the boundary with Book E1

Anchoring is the tendency of a number seen first to pull the judgment of numbers seen after it toward itself. Book E1, Chapter 7, calls the effect among the most replicated findings of decision research and cites Kahneman (2011); this book adopts that citation and adds a precision that matters for a pricing decision. What the evidence supports well is the direction of the effect and its persistence in laboratory judgments, including judgments made by people who know the anchor is arbitrary. What no source read for this book measures is its size on a business buyer facing a real quotation, with a budget, a procurement process and a colleague to convince. This book therefore uses anchoring as a reason to order the numbers on a page deliberately, not as a reason to expect a particular uplift. An owner promised a percentage by a consultant should ask which field experiment produced it, on what product, in what year.

The boundary with Book E1 is worth stating once. That chapter owns the sales conversation: how the price is presented, in what order, with what business case, and how it is defended against an objection. This chapter owns what is presented: where the anchor comes from, what level it sits at, and whether the comparison it invites is one the business can defend if a regulator asks. The two meet on the page, since the seller who opens with the buyer's cost of the problem is using a number this chapter's methods produced.

3.2.2 Where the anchor level comes from

An anchor is only as good as the number that sits in it, and there are three honest sources for that number in a business sale. The first is the buyer's own economics: the cost of the event the purchase prevents, the hours it saves, the premium it reduces. Chapter 2 builds that estimate and reaches an economic value to the customer of $5,668.00 a year for a forty-eight-unit rental building, of which the case captures 33.66% with an annual charge of $1,908.00. A page that begins with the buyer's exposure and then shows the annual charge anchors on a number the business can document.

The second is the alternative the buyer would otherwise buy. The case is compared with three competitors whose annual cost of ownership, kit amortized over five years plus twelve months of monitoring, is $1,368.00, $2,308.00 and $2,788.00. Anchoring on the highest would be convenient and would not be honest unless that competitor is genuinely the buyer's alternative; Chapter 2 selects the lowest, $1,368.00, because the selection has to survive a buyer who has done the same research. The third is the business's own published price: a list price that is really charged is an anchor the seller owns, and one that is never charged damages the seller the first time a buyer learns of the discount everyone else received.

There is a fourth source, and it is the one small businesses use by default: the seller's own cost. Chapter 2 shows the cost-plus price for the kit landing at $4,200.00 against a list price of $4,800.00 and an economic value far above both. The point to retain here is psychological rather than arithmetical. A seller who has spent six months building a product knows its cost intimately and the buyer's alternative vaguely, so the cost is the number that comes to mind first — and the number that comes to mind first is the seller's own anchor.

3.2.3 Anchoring against yourself

Three common practices anchor a business against itself, and all three appear in the case. The first is the old price. When Hydrosentinelle moved from a single subscription at $79.00 a month to a ladder of $89.00, $129.00 and $179.00, every existing customer had an anchor of $79.00 and every new customer had none. Chapter 8 treats the move on existing customers as an increase, with sixty days of notice and six months of grandfathering, and measures the tolerable loss at 28 subscriptions against the 6 the model assumes were lost. The anchor is not symmetrical: the existing customer compares $89.00 with $79.00, the new customer compares it with $129.00 and $179.00. The same number is read in two different comparison sets, and a business that publishes one page for both chooses the comparison for neither.

The second is the first quotation ever given to a buyer, which becomes that buyer's reference for every renewal; Chapter 7 treats the quotation as a document with a validity period for this reason among others. The third is the discount given early in a relationship, which anchors the buyer on the discounted price rather than on the list price, and which Chapter 6 measures. In all three cases the remedy is the same: decide the comparison before the buyer does, and write it down where the next person who quotes this buyer will see it.

3.3 Decoys and the Compromise Effect

3.3.1 What a third option does to a set of two

Add a third and more expensive option to a set of two, and something changes that has nothing to do with anyone buying the third option. The set that ran from "cheap" to "expensive" now runs from "cheap" to "middle" to "expensive", and the option that used to be the expensive one has become the reasonable one. That is the compromise effect, and the option added to produce it is called a decoy when no one is expected to buy it. Behavioral economics is one of the additions announced for the seventh edition of Nagle et al. (2023), and the mechanism is uncontroversial in the pricing literature; what this book does not have is a field measurement of its size, and Table 3.1 says so.

What a business can do, without any measurement, is compute what the top option is worth to its own average price under an explicit assumption about where its buyers would otherwise go. That computation is auditable, and it is honest as long as the assumption travels with the result. On the case, new buildings are assumed to divide 55% to Essential, 35% to Plus and 10% to Portfolio, giving an average revenue per building of $112.00 a month against the single old price of $79.00, an uplift of 41.77% that Chapter 5 attributes to the structure rather than to psychology.

3.3.2 What the top tier is worth on the case

Now withdraw the top tier and ask where its buyers go. The model assumes 70% of them fall back to Plus and 30% to Essential, which is deliberately conservative: it gives most Portfolio buyers to the next tier down rather than losing them. On that assumption, the average revenue per building falls from $112.00 to $105.80, a difference of $6.20 a building a month, or 5.86%. The interpretation matters as much as the number: the $6.20 is not what Portfolio buyers pay, but what the existence of Portfolio is worth across all buyers, including those who never consider it, through the choice it changes.

The table below sets the two ladders side by side. Its first column names the ladder, the next three give the share of new buildings assumed to choose each tier, and the last gives the average revenue per building those shares produce. Every share in the table is a model assumption, not an observation of the market.

Table 3.2 — The ladder with and without the top tier, on the case's assumed mix

Figure 3.1 brings the two halves of this chapter's arithmetic together: on the left, the same three prices said three ways, by the month, by the day and by the apartment of a forty-eight-unit building; on the right, the two bars of average revenue per building, $112.00 with the top tier and $105.80 without it, with the gap of $6.20 marked between them.

Figure 3.1 — One ladder, three ways of saying it, and what the top tier is worth. The left panel gives the three published prices per building per month ($89.00, $129.00 and $179.00), per building per day ($2.93, $4.24 and $5.88) and per apartment of a forty-eight-unit building ($1.85, $2.69 and $3.73), with the single earlier price of $79.00 a month, or $2.60 a day, below as the reference the first customers carry. The right panel gives the average revenue per building per month under the assumed mix: $112.00 with the three tiers and $105.80 if Portfolio is withdrawn and its buyers fall back 70% to Plus and 30% to Essential, a gap of $6.20 or 5.86%. Hydrosentinelle inc. is fictional, and the prices, the mix and the fallback are illustrative assumptions of the model.

3.3.3 The top option has to be real

Everything above depends on one condition, and it is legal and ethical at once: the top option has to be a real offer that a real buyer could want and could buy. On the case, Portfolio is real. It requires a minimum of 5 buildings, it carries a monthly service cost of $38.00 against $15.00 for Essential, and Chapter 5 sets out what it contains and why the difference is a fence rather than a decoration. A top option built the other way — priced high, containing nothing extra, never sold — is not a pricing device but a representation to the public about an offer that does not exist, and Section 3.8 explains which provisions it engages.

There is also a commercial reason, and it shows up in the numbers first. An option that is never chosen is one of the two symptoms that Book E1, Chapter 14, sends to this book, the other being discounting everywhere. A never-chosen option costs something every month: it takes space on the page, it takes time in every conversation, and it invites the buyer to wonder what is wrong with it. If it is still never chosen after a year of quoting, the honest conclusions are that the fence between it and the middle tier is invisible to buyers or that its price is wrong, and Chapter 5 tells you which.

WATCH OUT A decoy nobody could ever want is a representation about an offer that does not exist The temptation with a decoy is to build the cheapest possible one: a tier that exists only on the page, priced to make the middle tier look reasonable, with nothing behind it. Three things go wrong. First, the seller has made a representation to the public about a product, and if the general impression it conveys is false or misleading on a material point, the representation is reviewable conduct whether or not anyone was deceived, and a criminal offense if it was made knowingly or recklessly (Competition Act, ss. 52 and 74.01(1)(a)). Second, a buyer will eventually try to buy it, and a tier that cannot be delivered, or is delivered grudgingly at a loss, turns a paper device into an operational and reputational problem. Third, it stops working the moment one buyer compares notes with another, which in a market of property managers in one city takes about one meeting of the owners' association. The test before publishing a top option is worth writing into the price review: name the customer who would buy it, name what they get that the middle tier does not give them, and state the cost of serving them. If any of the three cannot be answered, the option is not an option.

METHOD Designing a three-option set that is honest and still works When to use it: whenever a price page, a proposal or a quotation is being built with more than one option, and again at each annual price review. Steps: (1) Decide the middle option first, because it is the one you want most buyers to choose. (2) Write the fence between the middle option and the one below it as a sentence a buyer could repeat, in terms of what the buyer gets rather than what you spend. (3) Build the top option for a named customer segment that exists, state what it contains that the middle does not, and compute its cost of service, as Chapter 5 does for the case at $38.00 a month against $15.00. (4) Build the bottom option so that it is genuinely usable and genuinely smaller; an entry option that fails the buyer produces a refund, not an upgrade. (5) Set the gaps between the three prices from the value and the cost of each tier, not from a ratio taken from a book. (6) Compute the average revenue per unit under your assumed mix, then recompute it with the top option withdrawn, and record both numbers with the mix assumption beside them, as Table 3.2 does. (7) Publish the three with the same amount of detail, so that the comparison is between offers rather than between amounts of ink. (8) After four quarters, measure how often each option is chosen, as Chapter 9 sets out, and rebuild any option chosen by nobody. Deliverable: a one-page option set giving, for each option, its contents, its fence, its price, its cost of service and its expected share, plus the two averages with and without the top option. Verification criterion: someone who did not build the set can name, for each option, the customer who would buy it and the one sentence that distinguishes it from its neighbor, and the average revenue computed from the expected shares matches the price list.

3.4 Charm Prices and Price Endings

3.4.1 What the field evidence actually shows

Prices ending in nine are the most studied device in this chapter and the only one for which this book can cite randomized field experiments. Anderson and Simester (2003) worked with two women's clothing mail-order catalogs from the same group, at an average price of about $50, randomizing customers by postal code and accounting for returns and cancellations. In the pilot, four dresses sold 66 units at prices ending in nine against 46 at five dollars less and 45 at five dollars more, roughly 40% more, while the ten-dollar spread between the two comparison prices changed almost nothing. The first full study estimated about 35%, the second about 15%, with new items gaining about 22% against about 10% for established ones. In the third, the effect was about 7% for items not already tagged as being on sale, and the "Sale" tag itself produced a larger increase than the nine ending.

Two findings matter more for a business audience than those percentages. The first concerns the mechanism: decomposing the price into its components, the authors found the two relevant coefficients significant but not different from each other, which does not support the popular explanation that buyers truncate the rightmost digits; what their data support is informational, a nine ending signalling that the price is a low one. The second follows from the first: when the information has already been delivered another way, the ending adds little, which is what the third study found where a "Sale" tag was present. The authors label that particular comparison a tentative finding worthy of further investigation, and this book repeats the caution rather than rounding it away. The paper also notes the context: the literature reports that between 30% and 65% of all prices end in nine.

Two cautions belong beside those results. The setting was a consumer catalog of clothing at about fifty dollars, sold to individuals who decide alone, while the case sells a monitoring subscription to a property manager who compares quotations and answers to an owner; nothing justifies transferring a percentage from the first setting to the second. And an informational signal has a direction: a nine ending signals a low price, which is precisely the signal a business positioned above its competitors on documented value may not want to send.

3.4.2 The arithmetic of an ending, on the case

The case gives a clean way to size the question, because the two ladders differ only in their endings. Rounded up to the next ten, the three tiers become $90, $130 and $180, and the average revenue per building under the same assumed mix rises from $112.00 to $113.00, which is $1.00 a building a month, or 0.89%. The interesting number is on the other side of the trade. The cost of serving an average new building is $22.90 a month and does not change with the ending, so the whole extra dollar is contribution: the average building contributes $112.00 less $22.90 under the published ladder and $113.00 less $22.90 under the rounded one. Measured on contribution rather than on revenue, the rounded ladder needs only 1.11% fewer buildings to give the entire gain back.

The table that follows states that trade in four lines. Its first column names the measure, the second gives the published ladder of $89, $129 and $179, the third the rounded ladder of $90, $130 and $180, and the fourth the difference. The last row is the one to act on: the gain is real, small, and erased by a volume loss far below anything this business could reliably measure.

Table 3.3 — The published ladder against the rounded ladder, on the case's assumed mix

This is the honest point of the section. A business of this size cannot detect a volume difference of 1.11%. Chapter 9 computes what detection would cost: seeing a lift of 5 percentage points on the win rate with ordinary confidence and power would take 1,477 quotations in each arm, 2,954 in total, which is 30.8 years at the company's volume of 96 a year. A device whose gain is smaller than the smallest effect the business can measure must be decided on other grounds, and the grounds available are positioning, consistency and the signal the ending sends.

CASE IN POINT Why the Plus tier is priced at 129 dollars and not 130 Fictional example; names and figures are invented. The decision took twenty minutes at the January price review and turned on three arguments rather than on an expected uplift. The first is in Table 3.3: the rounded ladder is worth $1.00 a building a month, 0.89%, and is erased by a volume loss of 1.11% the company cannot measure, so the arithmetic does not decide. The second was the signal. The cofounder responsible for sales pointed out that a ladder of $89, $129 and $179 reads as a considered ladder while $90, $130 and $180 reads as a rounded one; the president's objection, that a nine ending signals a discount rather than engineering, is the finding of Anderson and Simester (2003) read in the other direction, since an informational signal of a low price is not always welcome. The third argument decided it: a quotation for 12 buildings at the Plus tier is $1,548.00 a month, and the founders wanted a multiplication a property manager could check mentally, which is as true of $129 as of $130. They kept $129.00, recorded the decision and the three arguments in the pricing file, and wrote one line in the price review checklist: the endings are decided once a year, with the ladder, and never in a negotiation.

3.4.3 A rule for endings that a small business can hold

The practical rule that follows has three parts. First, choose an ending convention for the whole price list rather than for each price, because inconsistency is visible to buyers and creates work at every quotation. Second, let the convention follow the positioning: a business that competes on documented value and quotes to professional buyers gains little from a signal that says "this is cheap", while a business selling to consumers in a category where most prices end in nine pays a small price for standing out. Third, never let the ending be decided by the person who is afraid of the price, or in a negotiation, where a round number is an invitation to round further.

WHO DOES THIS The person who fears the price should not choose its last digits In a solo business the owner does everything, which makes the discipline harder rather than unnecessary: the owner writes the ladder on one day, with the value estimate and the cost of service in front of them, and fixes the endings on that same day, in writing, rather than in the moment before sending a quotation. In a small team, the price level belongs to the owner or the product lead, the endings belong to the same decision and the same document, and the seller applies them. At Hydrosentinelle the president owns the ladder with the product and engineering responsibility, the cofounder responsible for sales and operations owns the way it is presented, and neither changes a digit alone. The pattern to avoid appears in most small businesses without anyone deciding it: the person who feels the buyer's hesitation most keenly is the person who edits the number, and that person always edits it downward. In an established business the remedy is a separation of duties, which Chapter 11 treats: the person who sets the price is not the person whose pay depends on closing the deal. Chapter 6 gives the companion rule for the negotiation itself, where the authority to discount is bounded at 6% for a representative and 12% for the head of sales, and where a round number is never a reason.

3.5 Framing: Per-Day, Per-Unit, Partitioned and All-Inclusive

3.5.1 The same amount, said in another unit

Framing is the most available device in this chapter because it costs nothing and changes nothing: the amount owed is the same whichever unit expresses it. The case's middle tier is $129.00 a month, which is $4.24 a day, $2.69 a month for each apartment of a forty-eight-unit building, and $1,548.00 a month for a portfolio of twelve buildings. Each figure is exact, and each invites a different comparison: the monthly figure against other monthly bills, the daily figure against small discretionary spending, the per-apartment figure against the rent that apartment produces, and the portfolio figure against the budget of a manager who thinks in portfolios.

The table below gives the four framings for the whole ladder. Its first column names the way of saying the price, the next three give the amount for each tier, and the last says what the framing invites the buyer to compare it with. Every figure is the same three prices restated; the last column, not the numbers, is what the seller is choosing.

Table 3.4 — The same three prices, said four ways

Two remarks keep the table honest. The per-apartment figures divide by forty-eight because the reference building of Chapter 2 has forty-eight units, so a per-unit framing is only meaningful when the unit count is stated beside it. And the daily figures are the monthly price times twelve divided by three hundred and sixty-five, which is an average rather than an amount ever billed: nothing is charged daily, and the company's earlier single price of $79.00 a month was $2.60 a day by the same arithmetic.

3.5.2 Where a framing becomes a partial price

A framing becomes a legal problem at a precise point: when the number displayed most prominently is not the number the buyer must pay. That happens in two ways. The first is partitioning, where a price is shown as a base plus mandatory charges added later — installation, activation, an administration fee — so that the first price shown is not attainable. The second is installment framing, where a periodic amount is displayed prominently and the total faintly or not at all; a per-day figure printed twice the size of the monthly price belongs to this family.

Both are governed in Canada by provisions treated in Section 3.8, and the practical rule is simpler than the provisions. State the amount the buyer must actually pay, for the actual period of commitment, at least as prominently as any smaller number derived from it. A per-day figure beside a clearly displayed monthly price is a comparison aid; the same figure in place of the monthly price is a partial price representation waiting to be complained about.

DO THE MATH The same subscription framed four ways, and what none of them changes Fictional example; the prices, the mix and the reference building are illustrative assumptions of the model. Take the middle tier of the case. It is $129.00 a month for one building. By the day, it is $4.24. Per apartment of the forty-eight-unit reference building, it is $2.69 a month. For a portfolio of 12 buildings, it is $1,548.00 a month. Bought as a first year together with the installed kit, it is $6,348.00, the kit at $4,800.00 plus twelve months at $129.00. Five numbers, one obligation. Now the comparisons those framings invite: set against the earlier single price of $79.00 a month, or $2.60 a day, the daily framing makes the move look like a little over a dollar a day rather than $50.00 a month; set against the economic value Chapter 2 estimates at $5,668.00 a year, the annual framing of $1,908.00 makes the same price look like 33.66% of the value created. What none of the five numbers changes is what the buyer owes, what the business collects, or the contribution of $98.00 a month the Plus tier earns against its service cost of $31.00. Framing moves the comparison; it does not move the money. That is why it is cheap, and why it is regulated.

QUÉBEC & CANADA NOTE Advertising a price in Québec means advertising the total, with taxes apart Two regimes apply to a price advertised in Québec, and they do not cover the same buyers. Federally, the Competition Act applies to representations made to promote a product or a business interest, whether the buyer is a consumer or a business. Provincially, the Consumer Protection Act governs contracts between a merchant and a consumer; Hydrosentinelle sells to building owners and property managers, so most of its quotations fall outside the provincial regime, while a business that sells to individuals is squarely inside it. Under section 224 of that Act, as read for Book E1 on September 8, 2026, at the consolidation of April 7, 2026, a merchant may not, in an advertisement, give less prominence to the total price of a set of goods or services than to the price of a part of that set, may not disclose the amount of installments without disclosing and giving prominence to the total price, and may not charge a higher price than the advertised one (Consumer Protection Act, s. 224). The Office de la protection du consommateur, which enforces the Act, states the same requirement operationally: the advertised price must correspond to the total amount the customer has to pay, administration fees and transport and preparation fees are included in it, the Goods and Services Tax and the Québec Sales Tax may be excluded and added at payment, and the total price must be displayed more prominently than its components, the global price of a package predominating over the price of each element (Office de la protection du consommateur, 2025). The same page names as prohibited the advertising of periodic installments alone, charging more than the advertised price, and announcing a reduction that does not exist. Note the boundary of that citation: the Office is the enforcement body and a secondary source, cited as such; the numbered provisions on individual labeling were not read at the text of the Act for this book, and nothing is quoted from them. Verify the current text before relying on it.

3.5.3 All-inclusive is a decision about who carries the surprise

A business that publishes an all-inclusive price carries the variance itself. If installation runs long, the margin absorbs it; Chapter 6 shows exactly that in the case, where an unbilled installation overrun of $150.00 is one of the six lines that took 30.59% of the list margin of a deal. A business that partitions its price pushes the variance onto the buyer, who discovers it at the end. Between those two lies the option that is usually right for a small business: an all-inclusive price with a clearly stated scope and a named, priced option for what falls outside it. The scope is what makes the all-inclusive price defensible, and Chapter 7 treats it as a component of a quotation. Buyers who discover a mandatory charge at the end of a purchase do not experience it as a neutral rearrangement of the same total; they experience it as the seller having said something untrue. Section 3.7 treats what that costs in trust, and Section 3.8 what it costs in law.

3.6 Bundling and Unbundling

3.6.1 Why a bundle can be worth more than its parts

A bundle is one price for two or more things that could be sold separately. It can create value for both sides for reasons that have nothing to do with psychology: it reduces the number of decisions the buyer makes and the approvals the purchase needs, it reduces the seller's cost of selling and invoicing, and it smooths the differences in what individual buyers will pay for each component. It can also destroy value, in two ways that small businesses discover late. It hides the price of each component, so the seller stops knowing what each is worth and the buyer cannot tell what dropping one would save. And it almost always embeds a discount, because a bundle at the sum of its parts has no reason to exist in the buyer's eyes.

That last point is where the arithmetic starts. A bundle is a discount unless it is priced at the sum of the parts, and a discount has a break-even volume, exactly like the price decrease of Chapter 2. The question to ask of any bundle is therefore not whether buyers like it — they will — but how many more of them there must be to pay for the margin it gives away.

3.6.2 The case's first-year bundle

The case offers a natural bundle: the installed kit and twelve months of the Plus tier, sold together as a first year. Sold separately, the parts list at $6,348.00, the kit at $4,800.00 plus twelve months at $129.00. The bundle is priced at $5,999.00, a discount of 5.50%. The cost of the two parts together is $2,472.00, the kit at $2,100.00 plus twelve months of service at $31.00. Margin therefore falls from $3,876.00 to $3,527.00, a loss of 9% of the margin for 5.50% of the price — the ordinary asymmetry of any discount, and the reason Chapter 6 measures discounts in margin rather than in price. To break even, the bundle must sell 9.9% more first years than the parts would have sold separately.

The table below states that comparison line by line: the line, the parts sold separately, the bundle, and the difference. The last line is the decision criterion, and everything above it is arithmetic.

Table 3.5 — The case's first-year bundle against the same parts sold separately

Three observations follow. First, 9.9% is a modest hurdle, which is what makes bundling attractive when a business has reason to believe the bundle removes an obstacle to purchase — here, the separate approval that a capital purchase and a recurring subscription each require from a property owner. Second, a hurdle is not a certainty: the business has to observe whether first-year sales actually rose, which Chapter 9 explains how to do at a volume of 96 quotations a year and, just as importantly, when it cannot be done. Third, nothing in the bundle changes the cost, so every dollar of the discount comes out of margin, which is why the margin loss is nearly twice the price discount in percentage terms.

3.6.3 When to unbundle instead

Unbundling is the opposite decision and it is undervalued. It is the right move when the components have genuinely different buyers, when one is a recurring obligation and another a one-time purchase the buyer may want to delay, or when the bundled price has become an objection because the buyer believes they are paying for something they will not use. The case already unbundles in one place: additional sensor packs are sold on their own at $390.00 with a margin of $270.00, rather than folded into the kit, because buildings differ in how many sensors they need beyond the 24 the kit includes.

Unbundling also has a diagnostic use. If a business cannot state the margin of each component of its bundle, the bundle is hiding a loss, and the way to find it is to price the components as if they were sold separately. A bundle is a discount that nobody wrote down, in the same sense that an extended payment term is, and both belong in the margin model before they belong on the price page.

TOOLS OF THE TRADE What a price page and a quotation builder must let you control The devices in this chapter are decisions about what appears where, so a tool that cannot control placement is a tool that decides your pricing psychology for you. A price page builder must let you set the order of the options and mark one as recommended, so that the middle option is a deliberate choice rather than a consequence of alphabetical sorting. It must display the amount actually payable, for the actual billing period, more prominently than any derived figure such as a per-day or per-unit amount, and let you state the billing period in words beside the number. It must let you show a strike-through comparison price and store it with the dates during which it was the price charged, because a comparison price that is displayed but not recorded cannot be defended later; Section 3.7 explains why that record is the whole of the defense. It must show what each option contains in the same level of detail, so that the comparison is between offers. It must attach a bundle to its components and keep the components' own prices visible internally, so that the margin model of Chapter 2 can still compute the margin of each part. And it must end a promotional display on a date, automatically, rather than relying on someone remembering. Chapter 12 compares the categories of tool that do this work, without prices; the functions listed here are the requirements to bring to that comparison.

3.7 Reference Prices, Fairness and the Cost of Being Caught

3.7.1 How a reference price is formed

A reference price is the price a buyer carries in their head as the normal price for a thing. It is built from what the buyer paid last time, what the buyer has seen advertised, and what others who bought have said. A business influences all three, mostly without noticing: every quotation issued becomes a reference price for that buyer, every promotion run becomes one for the market, and every discount given to one property manager becomes, after one industry meeting, a reference price for several.

The most expensive version is the business that discounts routinely. After a year of it, the reference price in the market is the discounted price, and the list price is a fiction that serves only to make each discount feel like a concession. Chapter 9 measures the case on this point: across a year, 96 quotations were sent and 34 won, invoice realization was 94.21% of list value and pocket realization 91.21%, so roughly nine cents of every list dollar never arrived. A business that does not measure realization cannot tell whether its list price is still its price.

3.7.2 A was-price must be real and documented

The moment a business writes "was $6,348, now $5,999", it has made an ordinary-price claim, and the standard that applies is not the seller's good faith but a test with two branches. The Competition Act makes it reviewable conduct to represent the ordinary price of a product, whether the claim is about suppliers generally or about the advertiser's own price, unless a substantial volume was sold at that price or higher within a reasonable period, or the product was offered in good faith at that price or higher for a substantial period (Competition Act, s. 74.01(2) and (3)). The two branches are disjunctive, so either one suffices.

The numbers everybody quotes with those tests are not in the Act. The Competition Bureau's guidance supplies them: more than 50% of sales at the reference price or higher during the twelve months before or after the representation for the volume test, and the product offered in good faith at that price or higher for more than 50% of the six months before or after for the time test, with the period adjusted for seasonal products (Competition Bureau Canada, 2022). The same guidance explains that a claim about suppliers generally must rest on the prices competitors actually charge in the relevant geographic market, which for a small or medium-sized supplier is generally the municipality or the metropolitan area, and that a comparison failing both tests may still be lawful if it is not otherwise misleading, the example being a clearly identified liquidation with transparent disclosure of the original price.

For a business that also sells in the United States, the Federal Trade Commission's Guides Against Deceptive Pricing state the same principle in a different idiom: a former price must be the actual, bona fide price at which the article was offered to the public on a regular basis for a reasonably substantial period of time, and a price inflated to enable a reduction is fictitious (16 C.F.R. § 233.1). The canonical illustration is a pen ordinarily sold at $7.50, raised to $10 for a few days, then advertised as "Were $10, Now Only $7.50!", which the guides call a false claim because the advertised bargain is not genuine; the guides also warn sellers against implying that a former price was a selling price when it was only an asking price. Note the difference in force between the two American instruments used in this chapter: the part 233 guides interpret the prohibition on unfair and deceptive practices and create no civil penalty of their own, while the instrument discussed in Section 3.8 is a trade regulation rule carrying civil penalties.

The operational consequence is a record-keeping obligation that nobody enforces until it matters. Before publishing any was-price, write down the price, the dates it was charged or offered, and the evidence — invoices, the dated price page, the dated catalog — and keep it for at least the twelve months the volume test looks at. A business whose system displays a comparison price without storing it, which is the situation the In La Machine section below describes, has to keep that record somewhere else.

THE RULE An ordinary-price claim must pass the volume test or the time test Under the Competition Act, read on the Justice Laws Website, current to July 21, 2026, and last amended March 26, 2026, it is reviewable conduct to make, or permit to be made, a materially false or misleading representation to the public to promote a product or a business interest, and no proof of intent or of actual deception is required (Competition Act, s. 74.01(1)(a)). For price comparisons specifically, a representation as to the ordinary price of a product is reviewable conduct unless the suppliers of the relevant geographic market have sold a substantial volume of the product at that price or a higher one within a reasonable period, or have offered it in good faith at that price or a higher one for a substantial period; the same structure applies where the representation is clearly about the price at which the advertiser itself ordinarily supplies the product (Competition Act, s. 74.01(2) and (3)). A saving provision preserves a representation that was not otherwise materially false or misleading (Competition Act, s. 74.01(5)). Scope: federal law applying across Canada, in business-to-business sales as well as consumer sales. The thresholds of more than 50% of sales over twelve months and more than 50% of a six-month period come from the Competition Bureau's enforcement guidance, not from the text of the Act (Competition Bureau Canada, 2022), and should never be attributed to the statute. The same conduct, engaged in knowingly or recklessly, is a criminal offense under section 52. Verify the current text before relying on it.

3.7.3 What being caught costs

The sanctions attached to the civil provisions are not symbolic, and an owner who has never read them imagines something much smaller. On a first order under the reviewable-conduct provisions, an administrative monetary penalty can reach the greater of $750,000 and three times the benefit derived from the conduct for an individual, and the greater of $10,000,000 and three times that benefit for a corporation; where the benefit cannot be reasonably determined, the corporate cap becomes 3% of annual worldwide gross revenues, and the figures rise to $1,000,000 and $15,000,000 for subsequent orders. The tribunal may also order restitution to purchasers, and the stated purpose is to promote compliance rather than to punish (Competition Act, s. 74.1). The same provision holds the defense that makes a compliance routine worth running: no penalty is imposed on a person who establishes that they exercised due diligence to prevent the conduct.

The criminal side is harsher and requires more proof. Section 52 applies to a person who knowingly or recklessly makes a materially false or misleading representation to the public, without proof that anyone was deceived; on indictment, the fine is at the discretion of the court and imprisonment can reach fourteen years, while on summary conviction the maximum fine is $200,000 and imprisonment one year (Competition Act, s. 52). The determining test is the one already quoted: the general impression conveyed as well as the literal meaning (Competition Act, s. 52(4)). A display that is technically accurate and generally misleading is caught.

Enforcement is real and it is slow, which is its own lesson. The Competition Bureau's guidance on drip pricing records that the Bureau sued a national cinema chain in May 2023, succeeded before the Tribunal in September 2024, and that the matter was on appeal in January 2026; the page reports no amount, and this book prints none, because none was read at the source (Competition Bureau Canada, 2026). Three years of litigation over the way a fee was displayed is a cost that appears in no penalty table.

The cost that matters most to a small business, though, is none of these. It is the property manager who discovers that the "regular price" on the website was never charged. In a market of a few hundred buyers who meet at the same association, the reference price is set by conversation, and a business caught inflating one loses the anchor it spent years building.

3.8 Where Psychology Ends and the Law Begins

3.8.1 The line, and how a court finds it

The line between a device that informs and a device that misleads is drawn by a single test applied in two places. Federally, the court considers the general impression conveyed by a representation as well as its literal meaning (Competition Act, s. 52(4)); in Québec, the same approach governs consumer representations (Consumer Protection Act, ss. 218 and 219, as read for Book E1 on September 8, 2026). A seller therefore cannot defend a display by parsing it. "Starting at $89" printed large, above a line in small type saying that installation is mandatory and extra, is literally true and generally misleading, and the general impression is what decides.

The second thing to understand is the difference between the two federal tracks. Part VII.1 of the Competition Act creates reviewable conduct: no criminal record, no proof of intent, no need to show anyone was deceived, and the penalties of section 74.1. Section 52 creates a criminal offense and adds one element, the state of mind: knowingly or recklessly. The same display can fall on either side, which is why a business that discovers an error in a price display should correct it and document the correction immediately.

3.8.2 Partial price representations, and the exception for taxes

Since the 2022 amendments, Canadian law addresses the partial price display directly and in both tracks. It is a false or misleading representation to represent a price that is not attainable because of fixed obligatory charges, unless those charges are imposed only by or under an Act of Parliament or of a legislature, which is the exception for sales taxes (Competition Act, ss. 52(1.3) and 74.01(1.1)). The two provisions mirror each other, one criminal and one civil. The Competition Bureau's guidance states the position operationally: a business should never advertise a price that is not attainable, and the test is whether the consumer can rely on the advertised price as the price they will pay; charges entirely imposed by law are excluded, a variable rather than fixed charge may still raise concerns under the general provisions, and since the amendments the Bureau no longer has to demonstrate that the partial display was deceptive (Competition Bureau Canada, 2026).

This is the provision that governs the framing decisions of Section 3.5, and the reason a per-day figure is safe only when the payable amount is at least as prominent. It is also the provision behind Book B1's practical rule, which this book adopts: the price the buyer sees first is the price they pay, apart from taxes. For the case the application is concrete. A quotation for twelve buildings shows kits at $54,144.00, then the Goods and Services Tax at $2,707.20 and the Québec Sales Tax at $5,400.86, for $62,252.06; the taxes are stated separately because they are imposed by law, and everything else the buyer must pay is inside the first number. Chapter 7 builds that quotation line by line.

WHAT CHANGED In the United States, total-price disclosure became enforceable in May 2025 The Federal Trade Commission's Rule on Unfair or Deceptive Fees, published at 90 Federal Register 2166 on January 10, 2025, took effect on May 12, 2025 (16 C.F.R. pt. 464). Three things about it matter here. First, its scope is narrow: it covers tickets to live events and short-term lodging, not the rest of the economy, so a Canadian manufacturer of monitoring equipment is outside it. Second, its standard is strict where it applies: the total price is defined as the maximum total of all fees a consumer must pay for the goods or services and any mandatory ancillary good or service, excluding only government charges, shipping charges and fees for genuinely optional items, and shipping charges qualify only where they reasonably reflect what the business incurs; the total price must be disclosed more prominently than any other pricing information, and the nature, purpose and amount of any excluded charge must be disclosed before the consumer consents (16 C.F.R. §§ 464.1 and 464.2). Third, and this is why it belongs in a chapter on psychology, it prohibits no type or amount of fee, no itemization and no dynamic pricing: it regulates prominence, not price. That is the direction of travel on both sides of the border — the amount a seller charges remains the seller's decision, and the honesty of its display does not. Unlike the Guides Against Deceptive Pricing, which interpret the general prohibition on deceptive practices, this is a trade regulation rule and carries civil penalties. Verify the current text before relying on it.

3.8.3 The four checks before a price goes public

Selling above an advertised price is its own provision, and it is the one small businesses breach by accident. It is reviewable conduct to advertise a product for sale or rent in a market and then, during the period and in the market covered by the advertisement, to supply it at a higher price; the exceptions are narrow, covering a catalog that prominently states that prices may be in error where the advertiser proves the error, an advertisement immediately followed by a correction, securities, and supply by a person not in the business of dealing in the product (Competition Act, s. 74.05). A promotion that ends on the website but not in the quotation template is that breach, which is why Chapter 10 puts the end date of every promotion on the pricing calendar.

The table below is the check to run before any price, page, promotion or quotation goes out. Its first column names the device, the second states what keeps it lawful, the third states what makes it reviewable conduct or an offense, and the fourth names the provision. It is deliberately short: four lines are what a small business will actually run.

Table 3.6 — Four checks before a price goes public

Two devices are missing from the table because they raise no issue as long as the first line is respected: an ending digit and a per-unit framing say nothing false. That is the summary of the chapter's legal content. Psychology becomes law where a number is displayed that the buyer cannot actually pay, or a comparison is offered against a price that was never real.

IN LA MACHINE A promotion is a whole-number percentage and two dates In La Machine, a promotion (business/promotions, and the same page reached as offers/promotions) carries a name, a tagline, a description, an image, a whole-number percentage from 0 to 100, a status set by hand, a start date, an end date and a list of targeted items, which may be a subscription, a package or a product; sending it creates internal alerts and writes a simulated email line in a log, and the client side (client/promotions) shows the reduced price. On Business → Store → Products (business/shop, business/shop-product-edit), a product carries its price, its currency, its product cost, its shipping cost and fees, and a profit panel, and a strike-through comparison price can be displayed but is not saved. What is not here: because that comparison price is not stored, the platform keeps no record of what the ordinary price was or of the dates it was charged, so nothing in it can support the ordinary-price tests of Section 3.7, and that evidence has to live somewhere else. There is no promotional code, no fixed-amount discount, no usage limit, no minimum order, no stacking rule, no free month and no credit; a promotion is not applied at payment, the checkout being a demonstration (/checkout-demo) and /client-subscribe being disabled. There is no tier field on a package, no price-page builder, no control over the prominence of one number against another, and no check of any price presentation against the Competition Act or the Consumer Protection Act, which Book B1, Chapter 6, states already and this book confirms. Routes verified in the route inventory on September 19, 2026, and behaviors read in the application's source at its commit of September 7, 2026 (La Machine, 2026).

Costly Mistakes

• Buying a percentage of uplift from a consultant who cannot name the field experiment behind it. The only randomized field evidence cited in this chapter concerns nine endings in two clothing catalogs, and its authors report results from about 7% to about 35% by context.
• Putting a top option on the page that nobody could buy. A tier with no contents, no named buyer and no cost of service is a representation about an offer that does not exist, judged on the general impression it conveys rather than on the seller's intention.
• Printing a per-day figure larger than the amount actually payable. The arithmetic is exact and the display is a partial price representation, reviewable conduct federally and, for consumer sales in Québec, a practice the enforcement body names explicitly.
• Writing "was $6,348" without the records that prove it. A comparison price must have been charged in substantial volume, or offered in good faith for a substantial period, and a screen that displays a strike-through price without storing it proves nothing.
• Deciding the ending digits during a negotiation. The ending belongs to the annual price review, with the ladder; decided in the moment, by the person who feels the buyer's hesitation, it always moves downward.
• Spending an afternoon on psychology and none on the waterfall. These devices move tenths of a point, while Chapter 6 shows a single deal in which 30.59% of the list margin disappeared in six lines that nobody negotiated.

Key Takeaways

• A buyer judges a price only by comparison, so the seller's real decision is what the price is compared with, and a price published alone delegates that decision to the buyer.
• The evidence is thinner than the literature suggests: one field measurement is cited here, on nine endings in consumer catalogs, and the rest is mechanism plus arithmetic, as Table 3.1 states row by row.
• Anchoring is worth using and not worth quantifying: the direction is well established, and no source read for this book measures the size of an anchor on a business price.
• The top tier is worth $6.20 a building a month on the case, or 5.86% of the average revenue per building, under a stated assumption about where its buyers would otherwise go.
• A nine ending is worth 0.89% on the case and is erased by a volume loss of 1.11%, far below what a business sending 96 quotations a year could detect.
• Framing changes the comparison and never the money: $129 a month, $4.24 a day and $2.69 per apartment are one obligation, and the payable amount must stay the most prominent number.
• A bundle is a discount: the case's first year at $5,999 gives up 5.50% of the price and 9% of the margin, and needs 9.9% more first years to pay for itself.
• Psychology becomes law at two points: a first price the buyer cannot actually pay, and a comparison price that was never real, with penalties reaching the greater of $10,000,000 and three times the benefit for a corporation on a first order.

Your Action Plan

1. Write down what each of your prices is currently compared with. The deliverable is a one-page list of your prices with, beside each, the comparison a buyer most likely makes today and the comparison you would prefer; it takes an hour, and it is done when every price has a named comparison rather than a blank.
2. Build one honest anchor from your buyer's economics. The deliverable is a short statement of what the buyer gains or avoids in a year, with the source of each figure, placed above your price on the page and in the proposal; it takes two hours with the value estimate of Chapter 2, and it is done when the anchor rests on a document rather than an opinion.
3. Audit your top option. The deliverable is a written answer to three questions — who buys it, what they get that the tier below does not give them, and what it costs to serve — plus the average revenue per unit with and without it; it takes an hour, and it is done when all three are answered or the option is rebuilt.
4. Decide your ending convention once, in writing. The deliverable is a one-line rule in your pricing file stating how every price on your list ends and who may change it; it takes fifteen minutes, and it is done when the rule is dated and no price on the current list breaks it.
5. Check every price display for prominence. The deliverable is a pass over your price page, your quotation template and your advertisements, confirming that the amount actually payable for the actual period is the most prominent number and that every mandatory charge is inside it; it takes two hours, and it is done when no derived figure is more prominent than the payable amount.
6. Compute the break-even volume of every bundle you offer. The deliverable is, for each bundle, the list value of the parts, the bundle price, the discount in price, the loss in margin and the extra volume needed to break even; it takes an hour with your margin model, and it is done when each bundle carries its hurdle beside it.
7. Open a was-price file. The deliverable is a folder holding, for every comparison price you publish, the amount, the dates it was charged or offered, and dated invoices or price pages as evidence; it takes an hour to open and five minutes per promotion afterwards, and it is done when you could answer the volume test and the time test for any comparison price you display.

In La Machine

Every route in this section was verified in the route inventory on September 19, 2026, and the behaviors were read in the application's source at its commit of September 7, 2026 (La Machine, 2026); Book N8 will cover these pages one by one, and this section only describes what the pricing devices of this chapter can and cannot be expressed with.

• The promotions page (business/promotions, and the same page reached as offers/promotions) holds a name, a tagline, a description, an image, a whole-number percentage from 0 to 100, a status set by hand, a start date, an end date and a list of targeted items, and the client view (client/promotions) displays the reduced price.
• Sending a promotion creates internal alerts and writes a simulated email line in a log, so the sending is a record rather than a delivery, and the start and end dates are the one genuine control the platform gives over a promotional period.
• Business → Store → Products (business/shop, business/shop-product-edit) holds the price, the currency, the product cost, the shipping cost and fees, and shows a profit panel; a strike-through comparison price can be displayed, and it is not saved.
• Business → Session Packages (business/packages, productivity/packages) and the subscription packages page (business/subscription-packages) each hold a price and a cost and compute profit per sale and margin on screen without storing them, which is what makes a margin visible for each option of a ladder built by hand.
• What is not here. There is no stored comparison price and therefore no record of an ordinary price or of the dates it was charged; no promotional code, no fixed-amount discount, no usage limit, no minimum order, no stacking rule, no free month and no credit; a promotion is never applied at payment, the checkout being a demonstration (/checkout-demo) and /client-subscribe being disabled; and there is no tier field on a package, no control over which number is displayed most prominently, no bundle object, and no check of any price presentation against the Competition Act or the Consumer Protection Act, which Book B1, Chapter 6, states already and this chapter confirms.

Self-Check

1. Why can a buyer not judge a price without a comparison, and what are the three honest sources of an anchor in a business sale?
2. What is the top tier of the case worth to the average revenue per building, and what assumption is that figure conditional on?
3. What did Anderson and Simester actually measure about nine endings, and what does their evidence not support?
4. At what point does a per-day framing become a partial price representation, and which provisions govern it?
5. What must be true of a was-price before it is published, and where do the thresholds of 50% over twelve months and 50% of six months come from?

Answers. (1) Because a price carries no verdict on its own: the buyer compares it with the last price paid, the alternative, the budget or the cost of the problem, and will choose a comparison if the seller does not offer one; the three honest anchors are the buyer's own economics, the alternative the buyer would otherwise buy, and the business's own genuinely charged list price (Sections 3.1.1 and 3.2.2). (2) It is worth $6.20 a building a month, or 5.86%, taking the average from $112.00 to $105.80 if it is withdrawn, conditional on the assumption that its buyers fall back 70% to Plus and 30% to Essential, which is a model assumption and not a measurement (Section 3.3.2). (3) They measured units sold in two women's clothing mail-order catalogs at about $50, randomized by postal code, and found increases of about 35%, about 15% and about 7% across three studies, larger for new items and smaller when a "Sale" tag was already present; their data do not support the truncation explanation, and nothing in them supports transferring a percentage to a business-to-business subscription (Section 3.4.1). (4) At the point where the amount actually payable is no longer at least as prominent as the derived figure, so that the first price shown is not attainable; federally, sections 52(1.3) and 74.01(1.1) of the Competition Act, and for consumer sales in Québec, section 224 of the Consumer Protection Act (Sections 3.5.2 and 3.8.2). (5) A substantial volume must have been sold at that price or higher within a reasonable period, or the product offered in good faith at that price or higher for a substantial period, and the records must exist; the numerical thresholds come from the Competition Bureau's guidance, not from the Act (Section 3.7.2).

Chapter References

Anderson, E. T., & Simester, D. I. (2003). Effects of $9 price endings on retail sales: Evidence from field experiments. Quantitative Marketing and Economics, 1(1), 93-110. https://doi.org/10.1023/A:1023581927405 (full text read September 19, 2026, in the copy deposited by the author at the Kellogg School of Management, Northwestern University)

Bijmolt, T. H. A., Van Heerde, H. J., & Pieters, R. G. M. (2005). New empirical generalizations on the determinants of price elasticity. Journal of Marketing Research, 42(2), 141-156. https://doi.org/10.1509/jmkr.42.2.141.62296 (abstract and metadata read September 19, 2026, at the University of Groningen research portal)

Competition Bureau Canada. (2022, June 23). Ordinary price claims. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/how-we-foster-competition/education-and-outreach/publications/ordinary-price-claims

Competition Bureau Canada. (2026, August 24). Drip pricing. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/en/deceptive-marketing-practices/drip-pricing

Competition Act, RSC 1985, c C-34, ss. 52, 74.01, 74.05 and 74.1 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Consumer Protection Act, CQLR c P-40.1, ss. 218, 219 and 224 (Québec). LégisQuébec, consolidation of April 7, 2026, as read for Book E1 on September 8, 2026; not read again, LégisQuébec returning an access error on September 19, 2026.

Guides Against Deceptive Pricing, 16 C.F.R. pt. 233 (2026) (United States). Canonical source https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-233; text read September 19, 2026, at the Legal Information Institute, Cornell Law School, which publishes no consolidation date.

Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Marn, M. V., & Rosiello, R. L. (1992, September–October). Managing price, gaining profit. Harvard Business Review. https://hbr.org/1992/09/managing-price-gaining-profit (article page and publisher product description read September 19, 2026; the full text was not read, and the pagination reported by secondary sources could not be verified at source)

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Office de la protection du consommateur. (2025, February 10). Annonce d'un prix tout inclus. Gouvernement du Québec. Retrieved September 19, 2026, from https://www.opc.gouv.qc.ca/commercant/pratique-commerce/publicite-loi/prix-tout-inclus

Rule on Unfair or Deceptive Fees, 16 C.F.R. pt. 464 (2026) (United States) (source: 90 Fed. Reg. 2166, January 10, 2025; effective May 12, 2025). Canonical source https://www.ecfr.gov/current/title-16/chapter-I/subchapter-D/part-464; text read September 19, 2026, at the Legal Information Institute, Cornell Law School, which publishes no consolidation date.

Chapter 4 — Revenue Models

The question that opens this chapter arrived at Hydrosentinelle inc. in the form of a refusal. A property management firm with eleven buildings told Karim Haddad that the technology was fine, the references were fine and the price was fine, and that the company would still not be buying: the firm had no capital budget left for the year and would not create one for water sensors. Nothing about the offer needed to change. What needed to change was the shape of the exchange — when money moves, how often, what it is attached to and who carries the cost of the hardware in the meantime. Hydrosentinelle had exactly one shape, the one it had used since its first sale: an installed kit at $4,800 against a cost of $2,100, and a monitoring subscription at $79 a month against a cost of $15. That shape is a decision, and like the price level of Chapter 2 and the perception of Chapter 3, it had never been decided on purpose. This chapter treats the revenue model as the second of the four pricing decisions the book separates. It defines what a revenue model is and why it is settled before the price; it works through the one-time sale, recurring revenue, usage-based and hybrid models, the free first step, licensing and commissions, and outcome pricing; and it ends by putting two models on one offer over sixty months, measured in cash, in present value and in the size of the hole one of them digs before it pays.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Name the revenue model you are actually using, distinguish it from your price level and from your price structure, and say which of the seven models in this chapter your offer could also be sold under.
• Say who carries which risk under each model, so that a change of model is discussed as a transfer of risk and of financing rather than as a marketing idea.
• Compute what an annual prepayment discount really costs you, expressed against your own cost of capital rather than as a percentage that sounds small.
• Choose a usage metric that survives contact with a customer, using four tests that eliminate most of the metrics a business invents first.
• Price a free first step honestly, by costing the free users you will carry every month and by naming in advance the conversion you will measure.
• Compare two revenue models on the same offer in cash, in present value and in the depth of the cash trough, and say what financing the better model requires before it is adopted.

Why It Matters

A revenue model is the decision that determines whether a business can survive its own success. Two companies can sell the same thing, at the same total price, to the same buyer, and end the year in entirely different financial positions, because one collected in advance and the other collected over thirty-six months. The revenue model decides when cash arrives, which is the variable that closes businesses; how much working capital growth consumes, which is the variable that surprises them; and what a customer relationship is worth after the first sale, which is the variable that makes a business worth buying. Book A5 (Financing a Business) treats how a company funds itself and Book G2 (Cash Flow Management and Forecasting) treats how it manages the gap between paying and being paid; this chapter treats the decision that creates the gap, because that decision belongs to pricing.

For a solo owner, the revenue model is mostly a question of income stability. A business that sells only one-time engagements restarts at zero every January and spends a predictable share of every month selling rather than delivering; a retainer or a subscription converts part of that effort into a base that arrives whether or not the owner had a good month. For a small team, the model determines whether the team can be staffed at all, because a payroll is a recurring cost and a lumpy revenue line cannot carry it comfortably. For an established business, the model is a balance-sheet decision: moving from a sale to a service means financing the hardware, the software or the labor that the customer used to finance, and the lender, the accountant and the shareholders all have a legitimate interest in that decision before customers hear about it.

One boundary belongs at the outset. The revenue model is not the offer: what a tier contains, what it excludes and what guarantee accompanies it is the work of Book A6 (Designing a Value-Added Offer), and this book takes the offer as given. Nor is it the price structure, since how many tiers there are and how far apart they sit is the work of Chapter 5. This chapter decides the shape of the exchange; Chapter 5 divides it up, and Chapter 6 governs what may be given away inside it.

In This Chapter

• 4.1 What a Revenue Model Is, and Why It Is Chosen Before the Price
• 4.2 The One-Time Sale
• 4.3 Recurring Revenue: Subscriptions and Retainers
• 4.4 Usage-Based and Hybrid Models
• 4.5 Freemium, Trials and the Free First Step
• 4.6 Licensing, Royalties and Commissions
• 4.7 Performance-Based and Outcome Pricing
• 4.8 Comparing Two Models on One Offer: Sale Versus Hardware as a Service

4.1 What a Revenue Model Is, and Why It Is Chosen Before the Price

4.1.1 The shape of the exchange, not the size of it

A revenue model is the set of rules that say what a customer pays for, what triggers each payment, how often payments recur, how long the obligation runs on both sides, and who owns and finances the thing being used while it is being used. It is a different decision from the price level, which is how much, and from the price structure, which is how the price varies between buyers and between versions. A business can keep its price level unchanged, keep its structure unchanged, change only its revenue model, and transform its financial statements.

The distinction is easier to see on one object. Hydrosentinelle's kit can be sold outright for $4,800; it can be included at no separate charge in a thirty-six-month service commitment; it can be licensed to a security-alarm manufacturer that installs its own; it can be sold by a plumbing contractor who keeps a commission; or it can be installed free in exchange for a share of the water damage the building subsequently avoids. In every one of those sentences the hardware is identical and the customer's problem is identical. What differs is who pays, when, for how long, and who is holding the asset and the risk in between.

DEFINITION Revenue model, recurring revenue, annual contract value and churn A revenue model is the shape of the exchange between a seller and a buyer: what is charged for, what event triggers each payment, how often payments repeat, how long each party is committed, and who owns and finances the asset in use. It answers a different question from the price level, which is how much is charged, and from the price structure, which is how the charge varies across buyers and versions. Recurring revenue is revenue a business will collect again next period without making a new sale, because a live agreement produces it; a repeat purchase the customer must decide to make again is repeat revenue, and only the first of the two can be counted before it happens. Annual contract value is what one customer agreement produces over twelve months, before taxes and before one-time charges, and it is the unit in which subscription businesses compare customers of different sizes and billing periods; monthly recurring revenue is the same measure per month. Churn is the rate at which recurring revenue or customers are lost over a period: customer churn counts departures, revenue churn counts dollars, and the two differ whenever the customers who leave are not the average customer. Net revenue retention compares what this period's revenue from last period's customers came to, expansions included and new customers excluded, against what those same customers produced before; above one hundred percent, the existing base grew by itself. In plain terms: the revenue model decides the shape of the money, recurring revenue is the part that arrives without a new decision, and churn is the rate at which that part drains away.

4.1.2 Why the model is settled before the price

Four consequences follow from the revenue model, and all four constrain the price rather than the other way around. The first is the cash profile: a model decides whether money arrives before delivery, at delivery or over years, and Section 4.8 shows the same offer producing a positive cash position in its first month under one model and a hole of $1,867 under the other. The second is the cost to serve, because a model that keeps the seller responsible for hardware, hosting, support or replacement carries a monthly cost a one-time sale does not, for as long as the obligation lasts. The third is the unit of account — a device, a building, a seat, a month, a gigabyte, a result — which governs how the customer budgets, how the invoice is checked and how the relationship grows. The fourth is the customer's own accounting, which is the reason the property manager refused: a purchase is a capital expenditure that must be approved, budgeted and depreciated, while a monthly service charge is an operating expense a manager can often approve alone.

A price set before those four things are settled is a price for an exchange that has not been defined. That is why the strategic pricing literature places structure and policy before the price level rather than after it; the seventh edition of Nagle, Müller and Gruyaert organizes its chapters in that order and adds consumption-based and outcome-based models as new material, which is itself a statement about where the field's attention has moved (Nagle et al., 2023).

4.1.3 Seven models, and who carries the risk in each

The table that follows sets out the seven revenue models this chapter treats. Its columns give the model, what the customer understands themselves to be buying, the event that triggers a payment, the party that carries the main risk, and the section that treats it. The column to read hardest is the fourth, because a change of revenue model is above all a transfer of risk from one party to the other, and the party receiving the risk should be paid for it.

Table 4.1 — Seven revenue models, and who carries the main risk under each

KEY CONCEPT A revenue model decides who carries the risk and who finances whom Every revenue model answers two questions that have nothing to do with the amount of money involved: which party bears the risk that the thing does not work, and which party finances the other in the meantime. A one-time sale puts both on the buyer, who pays first and owns an asset whose value depends on results that have not happened yet. A subscription moves the performance risk to the seller, who has to keep delivering for a fee already agreed, while the buyer finances nothing beyond the current month. A service model in which the seller supplies the hardware moves the financing to the seller as well, which is why it must be priced as financing and not only as service. Outcome pricing moves both risk and financing to the seller entirely, which is why it commands the highest price and why so few small businesses can survive offering it. The practical consequence is a rule of negotiation: whenever a customer asks to change the shape of the payment — a longer term, a smaller deposit, a monthly charge instead of a purchase, payment on results — the customer is asking the seller to take on risk or to lend money. Both are legitimate requests, and both have a price.

4.1.4 Choosing among them

Most businesses do not choose a revenue model; they inherit the one their industry uses and discover much later that a different one was available. The method below is the deliberate version of the choice, and it is written to be run in an afternoon with people who are not finance specialists.

METHOD Choosing a revenue model in seven questions When to use it: before setting a price for a new offer, and again whenever a lost deal is explained by the shape of the payment rather than by its size, as in the refusal that opens this chapter. Steps: (1) Write down what the customer consumes over a year in physical terms — devices, buildings, hours, users, transactions, events — and pick the one quantity that grows when the customer gets more value. (2) Ask how the buyer's own budget treats the purchase, separating a capital expenditure that needs approval and depreciation from an operating expense a manager may approve alone, because that distinction decides more deals than the total does. (3) Decide who owns the asset at the end of the term, since ownership carries replacement, insurance and obsolescence. (4) Compute the seller's monthly cost to serve under each candidate model, support and replacement included, for the full length of the obligation rather than the first year. (5) Lay out the monthly cash of each candidate for sixty months from the month the first cost is incurred, and record the deepest negative point and the month the cumulative position turns positive. (6) Multiply that deepest point by the units the business plans to sell in a year and compare the result with the financing actually available. (7) Choose the model the business can finance, price it so the party taking the risk is paid for it, and write down the term, the cancellation rules and the ownership of the asset at the end. Deliverable: a one-page comparison of two or three candidate revenue models for one offer, each with its sixty-month cash line, its deepest trough, its payback month and its five-year total. Verification criterion: the trough multiplied by the annual unit plan is smaller than the financing the business has committed and available, and every figure on the page traces to a cost or a price in the accounting records.

4.2 The One-Time Sale

4.2.1 What it is good at

The one-time sale is the oldest revenue model and it is not a primitive one. A customer pays once, takes delivery, and owns the result; the seller books the revenue, keeps the cash and carries no further obligation beyond warranty. On the case, the kit sells for $4,800 against a cost of $2,100, which is $2,700 of margin, or 56.25 percent of the price, and that margin arrives in full at installation rather than being spread across a term.

Four advantages follow, and they are the reasons this model should not be abandoned lightly. The cash arrives immediately, so growth funds itself instead of consuming working capital. The accounting is simple, which matters more in a small business than anyone admits, because a model the owner cannot reconcile is a model the owner will not trust. The obligation ends, so a bad customer relationship can end too. And the offer is easy to compare, which shortens a sale when the buyer already knows what the thing is.

The model also has a structural weakness owners feel before they can name it: the year restarts at zero. Hydrosentinelle plans 180 kits in its third year at $4,800, which is $864,000 of kit revenue, and none of it arrives again in the fourth year unless 180 more are sold. An installed base is an asset only if something is attached to it, which is what the subscription does.

4.2.2 The deposit, the payment terms and the two prices nobody writes down

Two decisions inside a one-time sale are pricing decisions even though they look administrative, and Book B1 (Marketing Fundamentals for Business Owners), Chapter 6, names both as signals without costing them. This book costs them.

The first is the deposit. Hydrosentinelle asks 30 percent on order, which on the twelve-building quotation of Chapter 7 is $16,243.20 received before a single sensor is installed. A deposit does three things at once: it funds the materials, so the business is not lending the customer the cost of its own inventory; it filters buyers, because a customer who will not commit thirty percent has often not obtained internal approval; and it makes the project real on the buyer's side, because money already transferred changes the psychology of cancellation. What a deposit costs is deals, since some buyers genuinely cannot pay in advance. The right answer is usually to keep the deposit and trade it, which is the rule of exchange developed in Chapter 6.

The second is the payment term, the one most businesses give away without recording it. Chapter 6 computes what thirty extra days cost on one negotiated deal: on $4,286.40, at the 12 percent cost of capital the case model assumes, an extension from thirty to sixty days costs $42.28. That is small on one deal and not small across a year, and it never appears anywhere as a discount. Payment terms are price; so is a deposit waived, and so is an unbilled installation overrun, which on that same deal cost $150. Chapter 6 assembles them into the pocket price waterfall, which ends at $3,974.12, or 82.79 percent of the list price.

4.2.3 Where the one-time sale stops working

A one-time sale fails in three situations, and the third is the one that opens this chapter. It fails when the customer cannot own the asset, because a tenant, a franchisee or a manager without a capital budget has no way to hold it. It fails when the value depends on continuing work by the seller, because a buyer who owns a device but not the monitoring behind it owns very little. And it fails when the purchase is large relative to the buyer's decision authority, because a number above an approval threshold summons a committee, a business case and a delay, while the same total spread over months may not.

CASE IN POINT The property manager who wanted no capital expenditure Fictional example; names and figures are invented. Gestion Rivemont, a fictional property management firm, manages eleven rental buildings in the Eastern Townships and had been in discussion with Hydrosentinelle inc. for two months. The firm's position, put to Karim Haddad in a fifteen-minute meeting, was that eleven kits at $4,800 amounted to $52,800 of capital expenditure, that the owners approved capital expenditure once a year in a budget already closed, and that the earliest possible purchase was therefore fourteen months away. The firm added that it could approve a monthly service charge that same week, because building operating expenses were the property manager's own authority up to a monthly ceiling. Nothing in the technical evaluation was unresolved and no price had been contested. Laurence Bergeron, the president, first read the refusal as a discount request and prepared one; Karim Haddad read it correctly as a request for a different revenue model. The two positions are worth stating precisely: a discount would have reduced the total the firm paid and changed nothing about the approval it could not obtain, while a service model left the total higher and made the decision approvable. Section 4.8 puts numbers on both sides of that exchange.

4.3 Recurring Revenue: Subscriptions and Retainers

4.3.1 The fourth pricing method, and what it actually promises

Book B1, Chapter 6, told the library that four pricing methods would be treated in depth in this book: cost-plus, value-based, competitive and subscription. Chapter 2 delivered the first three. The subscription is the fourth, and it belongs here rather than there because it is not a method of computing a number; it is a method of shaping an exchange, and the number inside it is computed by the methods of Chapter 2 like any other.

A subscription is an agreement under which a customer pays a fixed amount per period for continuing access, availability or service, and continues paying until one party ends it. A retainer is the same instrument sold by a professional firm, where what recurs is reserved capacity — a block of hours, a response time, an availability — rather than access to a product. Both promise the seller revenue that arrives without a new sale, and both promise the buyer a predictable charge and the right to stop. The second half is the part sellers forget: recurring revenue is not contracted revenue unless a term says so, and a monthly subscription cancellable at any time is a forecast, not an asset.

On the case, the subscription is $79 a month against a cost to serve of $15, which leaves $64 of margin, or 81.01 percent. The company enters its third year with 206 buildings under subscription, plans 180 new installations, and loses 30 subscriptions to a 10 percent annual churn assumption, ending the year with 356 and averaging 281. Those 281 average subscriptions produce $266,388 of the year's $1,130,388 of planned revenue — nearly a quarter of the year decided in earlier years, arriving whether or not the sales year goes well.

4.3.2 What recurring revenue is worth, and what it costs

Three properties make recurring revenue valuable, and each has a cost attached that businesses discover late. Recurring revenue is predictable, which is what makes hiring, borrowing and planning possible; the cost is that the seller has committed a price for a period during which its own costs may rise, which is why Chapter 8 treats indexation and the increase at renewal as operations rather than hopes. It is cumulative, because each new customer adds to a base rather than replacing last month's; the cost is that the base must be served every month whether or not anyone bought anything, so a support failure now damages revenue already counted. And it is saleable, because a buyer of a business pays a multiple for a base that will still be there next year; the cost is that the same buyer will measure churn precisely, and a company that has never measured it will learn its number in a due diligence room.

Churn is the discipline that makes the other two properties real. A base growing by 180 installations a year and losing 10 percent of itself annually is a very different business from the same base losing a quarter of itself, and the difference compounds. Book G2 treats the cash consequences; this chapter's contribution is narrower: the price level and the billing period both move churn, and the evidence for that is now measurable.

4.3.3 What the measured data say about billing periods

Two published measurements bear directly on the choice between monthly and annual billing, and both must be read with their limits stated. They come from a vendor that measures subscription billing for its own customers, so the samples are non-probability samples of subscription software companies and not of businesses in general; neither report is peer reviewed; and the figures describe software, not services or hardware.

The first is a billing report published on May 18, 2025, covering the full year 2024, built from anonymized and aggregated billing data of more than 2,500 subscription software companies, excluding companies below $300,000 of annual recurring revenue and those relying entirely on a single billing model (Faustino, 2025). The second is a retention report published on December 10, 2025, covering 2025, built from approximately 3,500 vendor websites categorized with the assistance of artificial intelligence into roughly 2,700 business-to-business companies, 600 business-to-consumer companies and 200 artificial-intelligence-native companies, limited to companies that had reached $250,000 of annual recurring revenue (ChartMogul, 2025). The categorization method of the second report is itself a limitation worth naming, because a category assigned from a website is not a category verified with a company.

The table that follows gathers the four comparisons from those two reports that bear on the choice of billing period. Its columns give the segment measured, what the annual plans showed, what the monthly plans showed, and what the comparison means for a business deciding how to bill. Every row is a measurement of subscription software companies in a non-probability sample, and none of them is a rule about any particular business.

Table 4.2 — Annual and monthly billing compared in two non-probability samples of subscription software companies

Two further measurements from the same billing report are worth stating outside the table because they are about timing rather than mix. At low revenue per account, customer retention was 62 percent on annual plans against 41 percent on monthly plans, a gap of 21 points; and the second month of a customer's life was the peak moment for switching from monthly to annual billing, with customers reported as three times more likely to switch then than in the ninth month (Faustino, 2025). A business that offers an annual option and never mentions it again after the sale is missing the month in which the offer works.

4.3.4 What the retention measurements changed

WHAT CHANGED Measured subscription retention now separates sharply by price level, and one new category sits far below the rest Until recently, a small subscription business comparing itself with published benchmarks found a single band of retention figures and no way to tell whether its own price level explained its position. The retention report published on December 10, 2025 changed that in two ways, on annualized 2025 rates drawn from approximately 3,500 vendor websites categorized with the assistance of artificial intelligence, limited to companies past $250,000 of annual recurring revenue — a non-probability sample, categorized by machine, and not a survey of businesses at large. First, it separated a new category of artificial-intelligence-native companies whose median gross retention stood at 40 percent, against a business-to-business median net revenue retention of 82 percent with a top quartile at 97 percent, and a business-to-consumer median of 49 percent; the same category's median gross retention had been 27 percent in January of the same year, so the category is moving quickly and any figure attached to it dates fast. Second, and more useful to a reader of this book, it broke that category down by price level: products above $250 a month showed 70 percent gross and 85 percent net retention, products between $50 and $249 showed 45 percent and 61 percent, and products below $50 showed 23 percent and 32 percent (ChartMogul, 2025). The practical reading, as of September 19, 2026, is that a low price does not merely earn less per customer; in these measurements it attracted the customers who left. That is an argument for the price level of Chapter 2 and against the reflex of lowering a subscription to fill it.

4.3.5 The annual prepayment, priced as what it is

Book B1, Chapter 6, states that a discount for annual payment trades margin for cash and commitment, and sends the reader here for the arithmetic. Here it is, on the case.

Hydrosentinelle has always offered an annual prepayment at 10 percent off. Twelve months at $79 is $948 at list; the prepaid year costs $853.20; the discount is therefore $94.80. Stated that way it sounds like a rounding error. Stated as a financing transaction it does not, because that is what it is: the customer advances money the business would otherwise have collected month by month, and the business pays for the advance with margin.

The honest way to price it is against the company's own cost of capital, which the case model puts at 12 percent a year. Discounted at that rate, twelve monthly payments of $79 collected at the end of each month are worth $892.03 today. The prepayment brings in $853.20. The business is therefore paying $38.83 more than the money it receives is worth to it, on every prepaid subscription, every year. A 10 percent annual discount is not a 10 percent cost; it is a rate of borrowing well above what the business pays elsewhere, and the only reason to accept it is that it also buys something else. Section 6.5.3 reads the same trade the other way round, as an implicit annual rate on the cash actually brought forward; the two readings answer two different questions — whether to take the prepayment, and what rate is being paid for it — and they agree.

The table that follows sets the transaction out line by line so that an owner can repeat it with their own numbers. Its columns give the line, the amount, and what the line represents in the exchange. The line to look at is the last one, because a discount that costs more than the company's own cost of borrowing is a decision that has to be justified by retention rather than by cash.

Table 4.3 — What the annual prepayment costs, line by line, on one subscription

Three reasons justify keeping such a discount anyway, and each must be stated as a benefit rather than assumed: retention, because annual plans retained substantially better than monthly ones in the measurements above, and retention is worth more than the discount if the difference is real in your own base; collection cost, because one payment a year replaces twelve and the failed-payment rate of twelve charges is not zero; and certainty for the buyer, who may prefer one budgeted line to a charge that has to be re-justified. What is not a reason is that the prepayment "helps cash flow," said without a number, because that phrase conceals the fact that the business has just borrowed at a rate it would refuse from a lender. Book G2 manages the cash; this chapter prices the loan.

QUÉBEC & CANADA NOTE Advertising a monthly price in Québec, and what a prepaid year does to a small company's cash Two points bear on recurring revenue for a business operating in Québec, and they must be kept apart because only one of them is law. The first concerns display. Where the customer is a consumer, the Consumer Protection Act, CQLR c P-40.1, section 224, provides that a merchant may not, in an advertisement, disclose the amount of installments without also disclosing the total price and giving it prominence, may not give less prominence to the total price of an assembly of goods or services than to the price of a part of it, and may not charge a higher price than the advertised price. That is the reading made at source for Book E1 on September 8, 2026, at the consolidation of April 7, 2026; LégisQuébec returned an access error on September 19, 2026 and the text was not read again for this book. The Office de la protection du consommateur, which enforces the Act, states on its page of February 10, 2025 that the advertised price must correspond to the total amount the customer must pay, and that the GST, the QST and specific contributions remitted to a public authority may be added at payment (Office de la protection du consommateur, 2025). For a plan advertised at a monthly amount to consumers, the practical consequence is that an annual commitment presented as a monthly figure needs its yearly total shown at least as prominently. Hydrosentinelle sells to building owners and managers rather than to consumers, so the Act does not govern its quotations, but a reader selling a monthly plan to individuals is squarely inside it. The second point is not law: a prepaid year is a loan from the customer, and the arithmetic above shows what this business pays for it. A Québec company whose installation work concentrates in spring and autumn may reasonably pay for winter cash; the discipline is to decide that on purpose, price it against the operating line the business already has, and record the decision — which is the test Book A5 (Financing a Business) applies to every other source of money.

4.3.6 What has to exist before a subscription can be run

TOOLS OF THE TRADE What a subscription-billing capability must do, before any product is chosen A recurring revenue model is an administrative obligation before it is a growth story, and the obligation is the same whatever software is used. Recurring charge: a charge must be issued on a schedule without anyone remembering to issue it, and must survive a change of card, address or contact. Proration: a customer who starts, stops, upgrades or downgrades mid-period must be billed for what they had, with the rule visible to both sides. Failed-payment handling: a declined payment must be retried on a written schedule, escalated to a human at a defined point, and may suspend service only under a rule the customer was told in advance. Term and renewal: the system must hold a start date, a term, a notice period and a renewal date, and must report every agreement renewing in the next ninety days. Grandfathering: an existing customer must be able to keep an old price while new customers pay a new one, which is what makes the increase of Chapter 8 executable. Cancellation: a customer must be able to leave by a defined route and the date must be recorded, because a cancellation that lives in an inbox is a churn figure that does not exist. Measurement: monthly recurring revenue, churn and net revenue retention must be computable without a spreadsheet reconciliation. Export: the whole customer and billing history must come out in an open format, because a subscription base is the most valuable data such a business owns. Chapter 12 maps these functions against what a spreadsheet, a market tool and La Machine each cover, and the short version is that La Machine holds a monthly or annual period and nothing else in this list.

4.4 Usage-Based and Hybrid Models

4.4.1 What a usage metric has to survive

A usage-based model charges for a measured quantity: gigabytes stored, messages sent, transactions processed, kilometres driven, sensors monitored. Its appeal is alignment, because a customer who uses more pays more and a customer who is not yet sure can start small. Its difficulty is that most metrics a business invents fail at least one of four tests, and the failures show up months later as disputes. The first test is measurability by both parties, since a metric the customer cannot verify independently produces an invoice that gets queried every month. The second is alignment with value, meaning the metric rises when the customer's benefit rises. The third is predictability, because a buyer who cannot forecast next quarter's invoice within a reasonable band cannot get it approved. The fourth is the absence of a perverse incentive: a metric that punishes the customer for doing more of a good thing will be engineered around, and the seller loses the revenue and the goodwill together.

The table that follows applies those tests to four metrics Hydrosentinelle could plausibly use. Its columns give the candidate metric, what it counts, the condition under which it works, and the condition under which it fails. It is a worked example of the method rather than a recommendation, and the same four columns can be filled in for any business in an hour.

Table 4.4 — Four candidate usage metrics for the case, tested against value and verifiability

The fourth row is the interesting one. The metric most aligned with value is usually the one that cannot be measured to a standard both parties will sign, and a business that insists on it will spend its life arguing about a model rather than collecting an invoice. The workable answer for this case is the second metric, per sensor, combined with a per-building minimum — and the design of that combination belongs to Chapter 5, which builds the structure, while this chapter only establishes that usage is the shape of the exchange.

4.4.2 The per-seat model and its minimum

Per-user or per-seat pricing is a usage model with a very coarse meter, and it is the default in software for a reason: the count is unambiguous, the customer can verify it, and it grows when the customer's own organization grows. Book B1, Chapter 6, lists it among the seven structures a small business chooses from and attaches one condition to it that this book endorses without qualification: there must be a minimum. A per-seat price with no floor means the smallest customers pay an amount that does not cover the cost of serving them, and the cost of serving a customer is far less variable than the number of seats they buy. Onboarding a two-seat customer costs very nearly what onboarding a twenty-seat customer costs. The minimum is what makes the small end of the market survivable, and Chapter 5 sets where it sits.

4.4.3 Hybrid models, and why most businesses end up with one

A hybrid model combines a fixed recurring charge with a usage component: a platform fee plus metered consumption, a retainer plus hourly work beyond it, a base subscription plus a charge per additional unit. Hybrids exist because the pure models fail in opposite directions: pure usage-based pricing gives the seller no floor, so a quiet quarter at the customer's end becomes a bad quarter at the seller's, while pure subscription pricing gives the seller no upside, so a customer who triples consumption pays exactly what they paid before.

Hydrosentinelle is already hybrid and had not noticed: a one-time kit plus a monthly subscription is a fixed charge attached to an asset plus a recurring charge attached to a service. The observation matters because it reframes the question the company is actually facing. It is not choosing between a sale and a subscription; it already has both. It is deciding whether the asset should keep being sold, and that is the comparison of Section 4.8. Book B1's conclusion — that most small businesses end with a combination rather than a pure model — is chiefly a statement about structure, and Chapter 5 takes it up and puts numbers on the combination for this case.

4.5 Freemium, Trials and the Free First Step

4.5.1 Four different things that all look free

A free first step is not one model but four, and conflating them is the reason so many of them lose money. A free plan, the model usually called freemium, gives a permanently free version with reduced capability to anyone who wants it and expects a minority to convert. A time-limited trial gives the full product for a defined period and then stops. A paid pilot gives a limited deployment at a reduced price with a written decision point at the end. A demonstration gives a guided experience at the seller's cost and involves no customer deployment at all. The four differ in who does the work, how long the seller carries the cost, and — most importantly — in whether there is a date on which something must be decided.

Book B1, Chapter 2, sends this book the offers that sit in the middle of a customer journey, and these are three of them; the fourth, the renewal, is treated in Chapter 8. The pricing question for all of them is identical and is almost never asked: what does the free step cost per unit per month, and what conversion rate makes that cost worth paying?

The table that follows compares the four free first steps. Its columns give the form, what the buyer receives, what the seller carries while it runs, and the single measurement that proves whether it works. The last column is the one to write down before launching any of them, because a free step with no agreed measurement will be defended forever on the grounds that it generates goodwill.

Table 4.5 — Four forms of free first step, and the measurement that proves each one

4.5.2 The arithmetic nobody does before launching a free plan

WATCH OUT Free users cost money every month, and the bill does not stop when interest does A free plan is a recurring cost with no recurring revenue attached, and its cost is the seller's cost to serve multiplied by the number of free users, every month, for as long as they stay. On this case the arithmetic is brutally simple, because the cost to serve one building is $15 a month: a free monitoring tier would cost $180 a year per building, and a hundred free buildings would cost $18,000 a year before a single one converted. Three mistakes recur. The first is counting the cost only during a trial period, when a free plan has no end and the cost therefore has no end either. The second is forgetting that free users consume support, proportionally more of it than paying users because they are the least experienced part of the base. The third is treating the free base as a marketing asset without measuring whether it produces anything: a free plan is justified when free users convert at a rate that pays for all the free users, or when they bring paying users with them in a countable way, and in no other case. Before launching one, write down the conversion rate that would make it break even, then ask whether anything in the business's experience suggests that rate is attainable. If nobody knows, run a time-limited trial instead, which costs the same per user for a bounded number of months and produces the same evidence.

4.5.3 What may be said about a free offer, in two countries

The word "free" is regulated, and the regulation is more specific than most owners expect. The principle is common to Canada and the United States and fits in one sentence: an offer is free only if the customer pays nothing for the free item and no more than usual for whatever they must buy to get it. The detail differs by country and by the kind of instrument that carries it. In Canada there is no free-offer rule as such; a "free" claim is governed by the general prohibitions on false or misleading representations in the Competition Act, reviewable conduct under section 74.01 and, where made knowingly or recklessly, a criminal offense under section 52. In the United States the Federal Trade Commission has published a guide about the word itself, precise about frequency and about the meaning of the regular price; a guide is not a rule, since the Guide Concerning Use of the Word "Free" and Similar Representations interprets section 5 of the FTC Act and creates no civil penalty of its own.

The table that follows sets the two regimes side by side on four questions a business actually asks before running a free offer. Its columns give the question, the Canadian position, the United States position, and what the pair means in practice for a small business selling in both countries. Readers who sell only in Canada should still read the third column, because the American thresholds are the clearest published statement of what "regularly" and "occasionally" mean.

Table 4.6 — A free offer in Canada and in the United States, on four questions

The practical instruction for a small business is the same on both sides of the border, and Chapter 6 repeats it for promotions: a free offer needs a start date, an end date, a written statement of every condition disclosed at the outset rather than in a footnote, and a record of the ordinary price that was in effect before it. A free offer that never ends has quietly become the price.

4.6 Licensing, Royalties and Commissions

4.6.1 Licensing: selling the right to use rather than the thing

A license sells the right to use intellectual property — software, firmware, a design, a brand, a method — for a defined purpose, for a defined term, in a defined territory. It is available to any business whose value is partly in something reproducible, and small manufacturers routinely overlook it because they think of themselves as selling objects. Three variables define a license and they should be settled in this order. The term is perpetual or fixed; a perpetual license collects once and usually attaches a separate recurring maintenance charge, while a term license collects for as long as it is renewed, which makes it economically a subscription. The scope is what the licensee may do: one site, one product line, one country, one number of units. The base is what the payment is computed on, and it is where licenses go wrong.

A royalty is a license whose payment is a percentage of something the licensee does — units sold, revenue earned, savings generated. Its defect is that the base is defined by the other party's accounting, so the seller learns what happened only when the licensee reports it. Three protections make a royalty workable, and all three are ordinary practice: a minimum guarantee, so that the licensor is not wholly dependent on the licensee's effort; an explicit definition of the base saying whether it is gross or net of returns, freight, taxes and discounts; and an audit right with a threshold at which the licensee pays for the audit. Drafting these is the work of Book H1 (Contracts and Commercial Agreements); deciding that they are required, and what the minimum should be, is pricing.

For the case, the licensing option is real and narrow. Hydrosentinelle's value is partly in its firmware and its alert logic, which an established alarm manufacturer with a national installer network could deploy on its own hardware. A license would reach buildings the company will never reach itself, at a cost to serve close to zero, and it would put another brand on the company's intelligence. That trade-off is strategic rather than a pricing question; what pricing contributes is the observation that a license with no minimum guarantee gives the licensee the option to do nothing, and a licensee that does nothing has purchased the right to keep the licensor out of the market.

4.6.2 Commissions, referrals and the margin they come out of

A commission model pays a party for a transaction they facilitate but do not perform, and it runs in two directions, both of them pricing decisions.

Outbound, the business pays a commission to someone who brings it a sale — a plumbing contractor, an engineering firm, a property management association. That commission comes out of the same margin every other concession comes out of, and should be costed exactly like a discount, because that is what it is. A referral commission of ten percent on a $4,800 kit is $480 taken from a kit margin of $2,700, nearly a fifth of the margin on that sale, and the business should pay it only where the referral genuinely replaced selling effort. Two conditions make the arrangement sound: the commission is paid on collected revenue rather than on signed orders, so a bad debt is not paid for twice, and it is disclosed to the buyer whenever the referrer is in a position of advice, because an undisclosed payment to an adviser damages trust in both parties when it emerges, which it does.

Inbound, the business is paid a commission for transactions it facilitates for others, and that model raises the question that decides everything else about it: who is the customer? A business paid by a supplier for directing buyers to that supplier has the supplier as its customer, whatever the buyers believe, and the ethical and commercial consequences follow from that fact rather than from anybody's intentions.

One legal boundary belongs here even though its detail sits in Chapter 6. Where a business sells through resellers, the Competition Act treats influencing a reseller's price upward, or discouraging a reduction, as reviewable conduct under section 76 when it also has an adverse effect on competition in a market; the Act adds that suggesting a resale price is itself evidence of influence unless the supplier makes clear that the recipient is under no obligation to accept it and that its business relations will not suffer in any way. A business setting up a channel therefore decides its price, its discount to the reseller and the wording of any suggested retail price at the same moment.

4.7 Performance-Based and Outcome Pricing

4.7.1 The four conditions, and why most offers fail them

Outcome pricing charges for a result rather than for an input: a share of the savings generated, a fee per qualified lead, a percentage of a recovered claim, a payment triggered by a measured improvement. It is the model customers say they want and the model that fails most often, because it requires four conditions to hold simultaneously.

The result must be measurable by a method both parties agree on before the work starts, not reconstructed afterwards. It must be attributable to the seller rather than to the season, the economy or a change the buyer made at the same time. It must be within the seller's influence, because a seller paid for a result they cannot affect is gambling rather than pricing. And it must be bounded, with a cap that protects the buyer from paying an amount nobody anticipated and a floor that protects the seller from working for nothing. Book B1, Chapter 6, adds a fifth condition of its own, and this book adopts it: performance-based pricing should be combined with a base fee, so that the seller's fixed costs are covered while the variable part carries the incentive. A pure contingency model is a model in which the seller finances the buyer's experiment.

4.7.2 Why the case cannot use it, and what that teaches

Hydrosentinelle's value is unusually well suited to outcome pricing in principle and entirely unsuited to it in practice, and the reason is instructive. Chapter 2 measures what the company is worth to a forty-eight-unit rental building over a year: $8,900, built from the water damage avoided, the insurance premium credit and the hours the building's staff no longer spend, of which the company currently captures 33.66 percent. An outcome model would charge a share of that $8,900 instead of a subscription, and it cannot, for a reason that has nothing to do with the willingness of either party. The value rests on an assumption of half an event a year at a cost of $24,000, of which the system prevents 60 percent. Half an event a year means most buildings, in most years, have no event at all. There is no annual invoice to compute for a building where nothing happened, and no owner will accept a bill for a flood that did not occur on the grounds that it would have. Outcome pricing needs the law of large numbers, and the law of large numbers needs a portfolio: it becomes possible at the level of a property group with dozens of buildings, where the expected number of events is a budget line rather than a coin toss, and it remains impossible one building at a time.

The general lesson is the useful part: outcome pricing is a function of the frequency of the outcome, not of its size. A rare, valuable outcome makes a wonderful sales argument and a terrible invoice. A frequent, modest outcome — a lead, a transaction, a completed shift, a processed claim — makes a workable one.

THE RULE Competition Act, s. 74.01(1)(b): a claim about performance must rest on an adequate and proper test made before the claim is made Jurisdiction: Canada, federal. Instrument: Competition Act, RSC 1985, c C-34, paragraph 74.01(1)(b), in the consolidation current to July 21, 2026 and last amended March 26, 2026, read on the Justice Laws Website on September 19, 2026. The provision makes it reviewable conduct for a person, for the purpose of promoting a product or a business interest, to make a representation to the public in the form of a statement, warranty or guarantee of the performance, efficacy or length of life of a product that is not based on an adequate and proper test of it, and it places the burden of proving that the test was made on the person making the representation. Two consequences bear directly on a revenue model built on a promised outcome. First, the promise that supports the model — the percentage of damage avoided, the savings generated, the improvement delivered — is itself a performance representation, so the evidence for it has to exist before the offer is published rather than after a customer questions it. Second, the standard is the adequacy of the test rather than the sincerity of the belief, and a figure carried over from a supplier's brochure, a pilot with no control, or a single favorable customer result will not satisfy it. Part VII.1 of the Act is civil, so no intention needs to be proved; section 52 makes the same representation a criminal offense where it is made knowingly or recklessly. Scope for a reader of this book: every quantified claim of result, saving or durability that appears in an outcome-based offer, a price page, a quotation or a sales document. Verify the current text before relying on it.

4.7.3 Where it does work for a small business

Three conditions make outcome pricing workable at small scale, and a business that meets all three should consider it seriously. The outcome happens often enough to be billed monthly rather than annually. The measurement already exists in a system neither party controls, such as a payment processor, a public register or a utility meter. And the seller can survive a bad quarter, which in practice means the base fee covers the cost of delivery and the outcome component is upside rather than survival. Where those hold, outcome pricing captures more value than any other model, for the plain reason that it is the model in which the seller takes the most risk.

4.8 Comparing Two Models on One Offer: Sale Versus Hardware as a Service

4.8.1 The design of the alternative

The refusal that opens this chapter forced Hydrosentinelle to design a second revenue model for the same offer, and the design is deliberately minimal so that the comparison measures the model and nothing else. Under hardware as a service, the customer pays no kit fee. The company installs the same kit, retains ownership of it, and charges the same $79 monitoring subscription plus a monthly hardware supplement of $169 for a committed term of 36 months; after the term the supplement stops and the subscription continues at $79. Everything else — the sensors, the installation, the monitoring, the cost to serve — is identical to the sale.

The supplement is not a number someone liked. At the 12 percent annual cost of capital the case model assumes, the monthly payment that exactly recovers the $4,800 kit over 36 months is $159.43. The company charges $169, which is $9.57 a month more, and that difference is the price of four things the company now carries and did not carry before: the risk that the customer stops paying, the risk that the hardware fails and must be replaced at the company's expense, the cost of administering a thirty-six-month agreement, and the fact that the asset stays on the company's balance sheet. Over thirty-six months the supplement totals $6,084 against a kit price of $4,800.

4.8.2 The two models measured

DO THE MATH Sale versus hardware as a service over five years, on one building Fictional example; figures are invented and computed by the case model. One building, one kit, one subscription, sixty months, both models. Under the sale: month 0 produces the kit margin of $2,700 plus the subscription margin of $64, and every following month produces $64; over five years the company collects $6,540 of contribution, and the customer pays $9,540. Under hardware as a service: month 0 produces the subscription margin of $64 plus the supplement of $169 minus the $2,100 cost of the kit, which is −$1,867; months 1 to 35 produce $233 each; months 36 to 59 produce $64 each; over five years the company collects $7,824, and the customer pays $10,824. The difference in cash is $1,284 in the company's favor, and the customer pays exactly $1,284 more — the whole of the difference is the financing charge, because the hardware and the service are the same in both models. Discounted at 12 percent a year, the sale is worth $5,645.30 today and the service $6,027.35, a present-value advantage of $382.05 for the service. The cost of that advantage is the trough: the service model is $1,867 in the hole on the day of installation and does not return to zero until month 9, while the sale is positive from the first month. At the 180 kits the third year plans, that trough is $1,867 multiplied by 180, or $336,060 of working capital the sale model never needed.

The figure that follows, Figure 4.1, draws both models as cumulative cash lines over the sixty months, so that the trough, the payback month and the crossover are visible as shapes rather than as numbers in a table. The two five-year totals and the two present values are printed beneath the drawing, and the point of the picture is the distance below zero on the left, not the distance above the other line on the right.

Figure 4.1 — Cumulative cash of a one-time sale and of hardware as a service over sixty months. Two lines track the cumulative cash contribution of one building under two revenue models of the case. The one-time sale begins at $2,764, the kit margin of $2,700 plus one month of subscription margin, and rises by $64 a month to $6,540 at month 59. Hardware as a service begins at −$1,867, because the company absorbs the $2,100 cost of the kit and charges no kit fee, then rises by $233 a month — the $64 subscription margin plus the $169 supplement — turning positive in month 9, passing the sale line in month 28, and ending at $7,824 after the supplement stops at month 36. Discounted at 12 percent a year, the sale is worth $5,645.30 and the service $6,027.35; the customer pays $9,540 under the sale and $10,824 under the service. Hydrosentinelle inc. is fictional.

The table that follows repeats the comparison in figures and adds the two columns a drawing cannot carry: what each measure means and which party it favors. Its columns give the measure, the value under the sale, the value under the service, and the reading. The row to argue about is the trough, because it is the only one of the five that can end a company.

Table 4.7 — Sale and hardware as a service compared on one building over five years

4.8.3 What the comparison actually decides

Three readings come out of those five rows, in order. The service model is better on value and worse on cash, which is the general shape of every move from a sale to a subscription; that is not a paradox, since the company is lending the customer $4,800 at a rate above its own cost of capital, and lending is profitable and cash-hungry at the same time.

The trough is the constraint, not the profit. One building at −$1,867 is a rounding error; the third-year plan of 180 installations at the same trough is $336,060 of working capital that must come from somewhere before the first invoice pays it back. A company that converts without arranging that financing first will run out of money while its income statement improves, which is the most common way a profitable small business dies. Book A5 (Financing a Business) treats the instruments that fund it — an operating line, equipment financing against the very assets being deployed, a lender's view of a thirty-six-month receivable — and Book G2 (Cash Flow Management and Forecasting) treats the forecast that shows whether the arrangement holds month by month. The pricing decision and the financing decision are one decision, and this is where that becomes visible.

The customer's arithmetic is the reverse of the seller's. Gestion Rivemont pays $1,284 more over five years and receives a decision it can make this week within its own authority. Both parties are better off than they were at the refusal, and neither has conceded anything on price.

4.8.4 The decision rule

A business should move from a sale to a service model when four conditions hold together. Deals are being lost to the shape of the payment rather than to its size, which the business can only know if it records the reason for each loss. The financing exists and has been confirmed, in writing, for the trough multiplied by the annual plan. The hardware or the asset is reliable enough that the seller can carry replacement risk without a reserve that consumes the premium. And the term is enforceable, meaning there is a written commitment, a cancellation rule and a known remedy if the customer stops paying while holding the seller's asset. Where any one of the four fails, the correct move is a partial one: offer both models, price the service model at the supplement computed above, and let the customers who need it select themselves.

WHO DOES THIS The accountant and the lender must agree before a revenue model changes A change of revenue model is one of the few pricing decisions a business owner should not make alone, and the two people who must be consulted are rarely in the room. The accountant is consulted because the change alters revenue recognition, the balance sheet and the tax position: hardware that used to leave the books at sale now stays on them and is depreciated, revenue earned at delivery is now earned over a term, and the profit reported in the year of the switch falls even though the business has become more valuable. An owner who discovers that in the year-end statements will conclude the decision was wrong. The lender is consulted because the change consumes working capital, alters the assets available as security and may affect covenants written against ratios the new model moves mechanically. In a solo business both roles are bought for a few hours from an external accountant, and the conversation is worth more than the fee. In a small team such as this case, the president, Laurence Bergeron, owns the model decision and the head of sales and operations, Karim Haddad, owns the evidence that the market wants it; neither owns the financing, which is the constraint. In an established company the decision belongs on a finance agenda with a written comparison of the kind this section produced, reviewed again twelve months later against what actually happened. Chapter 11 develops these roles in full.

IN LA MACHINE Monthly or annual, and nothing in between A subscription package in La Machine (business/subscription-packages) holds a name, a price, a cost, a description, an image, an active flag and a billing period that may be monthly or annual and nothing else; the pair of price and cost is what lets the platform show a margin on screen. A session package (Business → Session Packages, business/packages and also reachable at productivity/packages) holds a name, a type of either meetings or minutes, a quantity, an unlimited option, a price, a cost and a currency, computes "Profit per sale" and "% margin" on screen without storing either, and credits the client's account; the page states that no payment is recorded. What is not here: no free trial and no trial period, no tier, no proration, no upgrade or downgrade path, no usage-based charge and no meter of any kind, no term, no notice period, no renewal date and no cancellation record, and no link at all between a subscription package and the real subscription of a real client. The client-side purchase path (client/packages and then /checkout-demo) displays a "DEMO MODE" banner and announces taxes "calculated at the next step" that are never calculated, and /client-subscribe is disabled. Nothing in the platform models a thirty-six-month commitment, a hardware supplement or an asset that stays on the seller's books, so the comparison of this section is built in a spreadsheet and the resulting decision is recorded in the platform as a price and a cost on a package.

Costly Mistakes

• Reading a request for a different payment shape as a request for a discount. A buyer without a capital budget is not asking for a lower total; on this case the service model left the customer paying $1,284 more over five years and was accepted, because it changed who could approve it.
• Switching to a recurring or service model without arranging the financing first. One building troughs at −$1,867 and recovers in month 9; one hundred and eighty of them trough at $336,060, and that money has to exist before the first invoice pays it back.
• Offering an annual prepayment discount without pricing it as a loan. Ten percent off a $948 year costs $94.80, and against the company's own 12 percent cost of capital it costs $38.83 more than financing the same cash by ordinary means.
• Launching a free plan without costing the free users. At a cost to serve of $15 a month, a hundred free buildings cost $18,000 a year, every year, and a free plan has no expiry date that stops the bill.
• Choosing a usage metric the customer cannot verify. A metric that produces an invoice the buyer cannot check produces a monthly dispute, and the most value-aligned metric is usually the least verifiable one.
• Promising an outcome the business cannot test. A quantified claim of performance must rest on an adequate and proper test made before the claim, and the burden of proving the test sits on the business making it.

Key Takeaways

• A revenue model is the shape of the exchange — what is charged for, what triggers payment, how often, for how long and who owns the asset — and it is settled before the price level, because it fixes the cash profile, the cost to serve, the unit of account and how the buyer's budget treats the purchase.
• Every revenue model is a transfer of risk and of financing, so a request for a longer term, a smaller deposit or a monthly charge instead of a purchase is a request for the seller to take risk or to lend money, and both have a price.
• A one-time sale is fast, simple and self-financing, and it fails when the buyer cannot own the asset, when the value depends on continuing work by the seller, or when the total exceeds the buyer's approval authority.
• Recurring revenue is predictable, cumulative and saleable, and each property has a cost: a price committed for a period, a base that must be served whatever the sales month looks like, and a churn figure anyone buying the business will measure precisely.
• An annual prepayment discount is a loan from the customer at a rate the seller sets. Twelve monthly payments of $79 are worth $892.03 today at the company's 12 percent cost of capital and the prepayment brings in $853.20, so the discount is justified by retention, not by the phrase "it helps cash flow."
• A usage metric must be verifiable by both parties, aligned with value, predictable enough to budget and free of perverse incentives, and a per-seat price needs a minimum because the cost of serving a customer varies far less than the number of seats bought.
• A free first step is four different things, and each needs its cost to serve computed per unit per month and its conversion measurement named before launch; an offer called free that never ends has become the ordinary price of the product.
• Moving from a sale to a service improved present value and destroyed short-term cash here: $6,027.35 against $5,645.30, bought with a −$1,867 trough per unit and a recovery in month 9, which is why the accountant and the lender decide this alongside the owner.

Your Action Plan

1. Write down the revenue model you are using today, in one paragraph. The deliverable is a paragraph naming what the customer pays for, what triggers each payment, how often, for how long, and who owns the asset in use; it takes about twenty minutes, and it is done when a colleague reads it and cannot find a payment event it does not cover.
2. Record the reason for every deal lost over the last twelve months, separating size from shape. The deliverable is a two-column list distinguishing losses on the amount from losses on the payment structure, the approval level or the ownership of the asset; it takes about two hours with your quotation records, and it is done when every lost deal has one of the two labels and the counts are written down.
3. Compute your cost to serve one unit for one month. The deliverable is a single number covering support, hosting, replacement, monitoring and anything else that recurs, traced to accounting records rather than estimated; it takes about half a day, and it is done when your accountant agrees the number is complete.
4. Price your annual prepayment against your own cost of capital. The deliverable is the five-line table of Table 4.3 filled in with your figures, ending with the amount the discount costs beyond financing the same cash normally; it takes about an hour, and it is done when you can state that excess in dollars per customer per year.
5. Test your candidate usage metric against the four tests. The deliverable is a four-column table like Table 4.4 listing every metric you could bill on, with the condition under which each works and fails; it takes about two hours, and it is done when one metric survives all four tests and a minimum has been proposed for it.
6. Build the sixty-month cash line of one alternative revenue model. The deliverable is a monthly cash series for one unit under your current model and under one alternative, with the trough, the payback month and the five-year total for each; it takes about half a day in a spreadsheet, and it is done when the trough multiplied by your annual unit plan has been compared with your available financing.
7. Put the change in front of your accountant and your lender before you announce it. The deliverable is a one-page comparison plus a written note of the revenue recognition, balance sheet and covenant consequences; it takes one meeting each, and it is done when both have confirmed in writing that the financing and the reporting consequences are acceptable.

In La Machine

Every route named in this section was verified in the route inventory on September 19, 2026 and read in the product's source code at commit be8bf04 of September 7, 2026 (La Machine, 2026); nothing below is inferred from a page title.

• business/subscription-packages holds a subscription package with a name, a price, a cost, a description, an image, an active flag and a billing period that may be monthly or annual only; the free-or-paid toggle on the page is stored in the browser rather than on the record.
• Business → Session Packages (business/packages, also reachable at productivity/packages) holds a package of meetings or minutes with a quantity, an unlimited option, a price, a cost and a currency, computes "Profit per sale" and "% margin" on screen without storing either, and credits the client's account; the page states that no payment is recorded.
• Account → Profile → Revenues & Expenses (business/subscriptions) reads subscription transactions, package purchases and product purchases and writes each sale as a revenue line with its cost lines beneath it, which makes it the only place in the platform where the margin of a recurring sale can be seen.
• client/packages leads to /checkout-demo, which displays a "DEMO MODE" banner and announces taxes calculated at a next step that never calculates them; /client-subscribe is disabled and says so.
• Control center → Plan tasks and Control center → Projects (chains) hold the steps of the action plan above as scheduled work with an owner and a due date, and Control center → Audit log records who changed a price or a cost on a package.
• What is not here. La Machine has no free trial and no trial period, no tier, no proration, no upgrade or downgrade path, no usage-based charge and no meter, no contract term, no notice period, no renewal date and no cancellation record, and no link between a subscription package and any real client subscription. There is no annual-versus-monthly comparison, no churn or retention measure, no annual contract value and no recurring revenue total anywhere in the platform, and /dashboard shows demonstration widgets with no revenue or margin indicator. Payment is a demonstration; nothing is charged. A reader who runs a subscription business will keep the model, the churn and the renewal calendar outside the platform and use its price and cost fields to record the result.

Self-Check

1. What is the difference between a revenue model, a price level and a price structure, and which of the three is settled first?
2. Why is an annual prepayment discount a loan, and what did it cost on the case beyond the company's own cost of capital?
3. What are the four tests a usage metric must pass, and which of them does a per-litre-of-water-saved metric fail?
4. Why can the case not use outcome pricing on a single building, and what would have to change for it to become possible?
5. On the five-year comparison, which model produced more cash and what did that advantage cost in the first months?

Answers. (1) The revenue model is the shape of the exchange — what is charged for, what triggers payment, how often, for how long and who owns the asset; the price level is how much is charged, and the price structure is how the charge varies across buyers and versions. The revenue model is settled first, because it fixes the cash profile, the cost to serve, the unit of account and the buyer's own accounting treatment (Sections 4.1.1 and 4.1.2). (2) Because the customer advances money the business would otherwise have collected month by month, and the discount is what the business pays for the advance: twelve monthly payments of $79 are worth $892.03 today at the company's 12 percent cost of capital while the prepayment brings in $853.20, so the discount costs $38.83 beyond financing the same cash normally, on top of the $94.80 of margin given up (Section 4.3.5 and Table 4.3). (3) The metric must be measurable by both parties, aligned with value, predictable enough to budget, and free of perverse incentives; a per-litre-of-water-saved metric fails measurability, since no estimation method exists that both parties can verify, even though it is the most value-aligned candidate (Section 4.4.1 and Table 4.4). (4) Because the value rests on half an event a year, so most buildings in most years have no event and nothing to invoice; outcome pricing needs frequency rather than size, and becomes possible only across a portfolio large enough for the expected number of events to be a budget line (Section 4.7.2). (5) Hardware as a service produced $7,824 of cash against $6,540, and $6,027.35 of present value against $5,645.30; the advantage cost a trough of −$1,867 per building in month 0 and a return to zero only in month 9, which at 180 planned installations is $336,060 of working capital (Section 4.8.2, Figure 4.1 and Table 4.7).

Chapter References

ChartMogul. (2025, December 10). The SaaS retention report: The AI churn wave. Retrieved September 19, 2026, from https://chartmogul.com/reports/saas-retention-the-ai-churn-wave/ (annualized 2025 rates from approximately 3,500 vendor websites categorized with the assistance of artificial intelligence; a non-probability sample)

Competition Act, RSC 1985, c C-34, ss. 52, 74.01 and 76 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Consumer Protection Act, CQLR c P-40.1, s. 224 (Québec). LégisQuébec, consolidation of April 7, 2026, as read for Book E1 on September 8, 2026; not read again, LégisQuébec returning an access error on September 19, 2026.

Faustino, S. (2025, May 18). SaaS billing report: Billing is the invisible force behind your growth. ChartMogul. Retrieved September 19, 2026, from https://chartmogul.com/reports/saas-billing-report/ (aggregated billing data from more than 2,500 subscription software companies for full-year 2024; a non-probability sample of one vendor's own customers, excluding companies below $300,000 of annual recurring revenue)

Guide Concerning Use of the Word "Free" and Similar Representations, 16 C.F.R. § 251.1 (2026) (United States). Canonical source https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-251; text read September 19, 2026, at the Legal Information Institute, Cornell Law School, which publishes no consolidation date.

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Office de la protection du consommateur. (2025, February 10). Annonce d'un prix tout inclus. Gouvernement du Québec. Retrieved September 19, 2026, from https://www.opc.gouv.qc.ca/commercant/pratique-commerce/publicite-loi/prix-tout-inclus

Chapter 5 — Price Structures and Tiers

For twenty-six months Hydrosentinelle inc. had one price for monitoring, and that price was $79.00 a month for one building, whoever the buyer was and whatever the building contained. A six-apartment triplex in Sherbrooke paid it, and a four-hundred-unit portfolio paid it four hundred times. When Karim Haddad, the cofounder responsible for sales and operations, laid a year of service tickets beside a year of invoices, the single price stopped looking simple and started looking expensive: it was giving the same revenue for very different amounts of work, and it was inviting the largest and most demanding buyers to pay the least per unit of value they took. Chapter 2 had already settled the level, by measuring what a building owner gains and what the alternatives charge. What had not been decided was the shape: what the price is charged per, how many prices there are, what separates one from the next, what a buyer who buys a lot pays, and what happens when work cannot be scoped in advance. That shape is the price structure, and it is a decision of its own. This chapter makes it: it separates structure from level, defines the fence that holds two prices apart, builds a three-tier ladder and prices it, chooses a price metric and sets a minimum, writes a volume table and exposes the cliff inside it, prices an add-on and measures its attach rate, decides where hourly work still belongs and where a fixed price must carry the risk of a drifting scope, and assembles all of it into one price architecture that fits on a page.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Separate the structure decision from the level decision, and say which of the two produced any change in your average revenue per customer.
• Write the fence that holds two of your prices apart, in one sentence a buyer could repeat, and test whether it is verifiable rather than merely asserted.
• Build and price a good-better-best ladder, setting the gaps from the value and the cost of each tier rather than from a ratio borrowed from a book.
• Choose a price metric among the candidates your business actually has, and set the minimum that keeps your smallest customers from costing more than they pay.
• Write a volume table that does not contain a cliff, and compute what each band costs you in contribution and what it must buy back in units.
• Decide which of your work is priced hourly, which is priced fixed, and what the risk of a drifting scope is worth, so that the fixed price carries a number rather than a hope.

Why It Matters

Structure is the pricing decision a small business is most likely to skip, because a single price feels honest and is easy to say. It is also the decision with the largest arithmetic behind it. On the case, moving from one price to three raised the average revenue per building from $79.00 to $112.00 a month, an uplift of 41.77%, and only $10.00 of that $33.00 came from charging more for the entry offer. The other $23.00 came from buyers who, given a larger option that was genuinely better for them, chose it. No customer was persuaded of anything and no cost was cut; the shape of the price list did the work.

For a solo founder, structure is the way out of the trap in which every customer is quoted the same number and the biggest customers are therefore the least profitable. A single hourly rate or a single monthly fee means that the buyer who consumes the most of a founder's finite week pays exactly what the buyer who consumes the least pays. For a small team, structure is what makes a price list teachable. Two people can quote consistently from three named tiers and a volume table; they cannot quote consistently from one price and a shared instinct about when to bend it. For an established business, structure is where the margin of the next five years is decided, because a badly built ladder is very hard to withdraw: tiers that buyers have learned, minimums that were never set and volume bands that were written once in a hurry all become commitments that a later price increase has to work around. Chapter 8 treats that later work, and it is much harder than getting the shape right the first time.

There is one more reason, and it is why Book E1 (The Sales Process: From Prospect to Signed Contract), Chapter 14, sends readers here. Two symptoms bring most owners to a book on pricing: discounting everywhere, which is a policy problem that Chapter 6 owns, and options that nobody chooses, which is a structure problem owned here. An option that nobody chooses is not bad luck; it is a badly built option, and Section 5.3 sets out the four ways it can be badly built and how to tell which one applies.

In This Chapter

• 5.1 Structure Versus Level: Two Decisions, Not One
• 5.2 The Fence: What Legitimately Separates One Price from Another
• 5.3 Good, Better, Best
• 5.4 Per-Seat, Per-Unit, Per-Site — and the Minimum
• 5.5 Volume and Tiered Discounts
• 5.6 Add-Ons, Options and the Attach Rate
• 5.7 Hourly, Daily and Fixed-Price Work
• 5.8 Assembling a Price Architecture

5.1 Structure Versus Level: Two Decisions, Not One

5.1.1 What a price structure is

A price level answers the question "how much?" A price structure answers four other questions, and it answers them before the level can be set sensibly. What is the price charged per — per hour, per building, per user, per transaction, per month, per project? How many prices are there? What separates one price from another, which is the subject of Section 5.2? And what changes a price automatically, without anybody negotiating: a volume threshold crossed, a term committed to, a minimum enforced, an option added?

Those four questions are logically prior to the level, because a level is a property of a structure: "seventy-nine dollars" is not a price until someone says seventy-nine dollars for what, counted how, for how long, and instead of what else. The strategic pricing literature organizes the work in that order — value first, then structure, then policy, then level, then communication — and the current edition of the standard text is built around a value cascade running through those stages (Nagle et al., 2023). This book follows the same order: Chapter 2 established what the offer is worth, this chapter builds the structure, Chapter 6 writes the policy, and Chapters 7 and 8 communicate and maintain what results.

The table that follows sets the four decisions side by side, so that a reader can name which one is under discussion the next time a meeting is called about "the price." Its columns give the decision, the question it answers, the object it produces, and the chapter that treats it; the third column is the one worth reading twice, because a business that cannot point to the object has not made the decision.

Table 5.1 — Four pricing decisions, and the object each one produces

5.1.2 The four forms of a fixed price

Book A1 (From Idea to Business Model), Chapter 7, distinguishes a fixed price from a dynamic one and names four forms the fixed price can take. A business that is building a structure is choosing among those four, and it is worth naming them explicitly because most small businesses use one of them without knowing that the other three exist.

• A list price is a single published amount for a defined offer, the same for every buyer who meets the published conditions. It is the simplest structure, it is the easiest to defend, and it is the one Hydrosentinelle had for twenty-six months.
• A price by feature or by tier offers the same underlying service in two or more versions at two or more prices, so that a buyer who needs less pays less and a buyer who needs more pays more. This is the good-better-best ladder of Section 5.3.
• A price by segment charges different prices to different kinds of buyer for what is substantially the same offer, on the strength of a characteristic that can be stated and verified. This is the fence of Section 5.2, and it is the form most likely to go wrong, because a segment price without a fence is simply a discount granted to whoever asks.
• A price by volume varies with the quantity bought, either on a single order or cumulatively over a period. This is the volume table of Section 5.5.

Those four are not exclusive. The architecture of Section 5.8 uses all four at once: a list price for the installed kit, a tiered price for the subscription, a segment fence reserving the top tier for portfolios, and a volume table for kits bought together. What matters is that each is a deliberate object with a written rule behind it.

5.1.3 The case: how much of the uplift was structure, and how much was level

The case makes the distinction measurable, because both decisions were taken in the same quarter. The old structure was a list price of $79.00 a month for one building. The new structure is three tiers, all per building and per month: Essential at $89.00, Plus at $129.00 and Portfolio at $179.00. The model assumes new buildings divide 55% to Essential, 35% to Plus and 10% to Portfolio, giving an average revenue per building of $112.00 and an uplift of 41.77% over the single earlier price.

Now separate the two decisions. Had the company changed nothing but the number, putting every new building on a single price of $89.00, the average would have been $89.00, and the gain would have been $10.00 a building a month. That is the level decision, and it is the same decision Chapter 8 takes for the two hundred and six buildings that were already under subscription. The remaining $23.00 — the distance from $89.00 to $112.00 — is produced entirely by the fact that buyers who wanted more were given somewhere to go. That is the structure decision. Of the $33.00 of total uplift, the level carries $10.00 and the structure carries $23.00, which is to say that roughly seven dollars in ten came from the shape of the list rather than from asking more for the same thing.

That proportion is not a law of nature and should not be carried to another business as a rule of thumb. What is general is the method: compute the average revenue per customer under your current structure, recompute it under the new structure with the entry price held fixed, and attribute the difference to the shape.

KEY CONCEPT A tier the buyer cannot tell apart from the next one is a discount Every additional price on a list is a promise that something real separates it from its neighbor. When the separation is real, stated before the price and verifiable after the sale, the lower price is a legitimate offer to a buyer who genuinely needs less. When it is absent, vague, or discovered only after the contract is signed, the lower price is not a tier at all: it is a published discount, granted to every buyer who asks for the cheaper line, including those who would happily have paid the higher one. The test costs nothing to run. Ask someone who did not build the list to read the two lines and say, in one sentence and without looking at the prices, who should buy each one. If that person cannot do it, no buyer will do it either, and the lower price has become the only price. This is why the fence of Section 5.2 is written before the prices are set rather than after, and why a tier whose only difference is the amount of money is not a structure but a leak.

5.2 The Fence: What Legitimately Separates One Price from Another

5.2.1 Why a fence is necessary at all

The moment a business publishes two prices for things a buyer might regard as similar, it has created an arbitrage. Every buyer who would have paid the higher price now has an incentive to find a way to qualify for the lower one, and buyers are resourceful. A price fence is the condition that makes the cheaper offer genuinely unattractive or genuinely unavailable to the buyer the expensive offer was written for. Without one, the average price of the business falls to the lowest price it publishes, minus whatever is negotiated off that.

A fence is not a trick and not a refusal to serve. It states a difference that already exists between two buyers or two offers. A good fence has three properties: it is stated before the price, so the buyer chooses rather than discovers; it is verifiable, so the seller can tell which side of it a buyer is on without an argument; and it is acceptable when read aloud, which is the test that eliminates most fences built purely to extract money.

DEFINITION Price fence, price metric, tier and price architecture A price fence is the stated and verifiable condition that separates one price from another and prevents a buyer for whom the higher price was written from obtaining the lower one; it may rest on a characteristic of the buyer, on the contents of the offer, on the quantity or term committed to, on the channel, on the time of purchase, or on the conditions of payment. A price metric is the unit the price is charged per — an hour, a user, a building, a transaction, a site, a month — and it is the most consequential structural choice a business makes, because it decides whose growth the seller's revenue follows. A tier is one named line on a price list, consisting of defined contents, a price, a fence separating it from its neighbors, and a cost of service the seller knows; a list of tiers ordered from smallest to largest is a ladder, and the familiar three-rung form is called good-better-best. A price architecture is the whole structure together: the metric, the tiers with their fences, the volume table, the add-ons, the minimum and the currency, written on one page with an effective date. In plain terms, the metric says what you count, the tier says what you sell, the fence says who may buy it, and the architecture is all of that on one sheet.

5.2.2 Six kinds of fence, and what each one demands

Fences come in a small number of recognizable forms, and a business usually has more of them available than it thinks. The table that follows names six and gives, for each, what separates the buyers, what the seller must verify, and how that fence typically fails. Its last column is the one to read before adopting a fence, because every form has a characteristic failure that is easy to predict and hard to repair afterwards.

Table 5.2 — Six kinds of price fence, what each verifies, and how each one fails

5.2.3 The fence on the case, and the one that was refused

Hydrosentinelle's ladder rests on two fences of different kinds. Between Essential and Plus the fence is an offer contents fence: the two tiers differ in what the monitoring service does and in what the company commits to when a sensor reports, and the difference is visible in the cost of service, $15.00 a month against $31.00. Between Plus and Portfolio the fence is a buyer characteristic fence: Portfolio is available only to a buyer with at least 5 buildings under subscription, and the count is verifiable from the company's own records.

What the founders designed for each tier — the contents, the exclusions and the commitments — is offer design, and Book A6 (Designing a Value-Added Offer) is the volume that does that work. This chapter takes the contents as given and does the pricing: the amount on each line, the gap between the lines, and the fence that stops a buyer from stepping down. The two disciplines meet at one number, the cost of service, which the offer design determines and the pricing must respect.

A third fence was proposed and refused, and the refusal is instructive. Karim Haddad suggested charging property managers more than owner-occupiers for identical service, on the reasoning that a manager bills the cost through. The proposal failed the read-aloud test at once, since no sentence describing that fence would be acceptable to a property manager hearing it, and it failed the verifiability test too, since nothing prevents a manager from signing in the owner's name. A fence that depends on the buyer not understanding it is a trap with a delay on it, and the delay ends the first time two buyers compare notes.

WHO DOES THIS The product owner decides what goes in a tier, and the seller never does Three roles meet at a price list and they must not be collapsed into one. The person who owns the offer — the product owner where there is one, the founder responsible for the product where there is not — decides what each tier contains and excludes, because that person alone knows what can be delivered at scale. The person who owns the numbers supplies the cost of service for each tier and must be willing to defend it, because a tier priced against a guessed cost is priced against nothing. The owner approves the list and signs the effective date. The seller is consulted, heard carefully and given no authority over the contents, for a reason that has nothing to do with trust: a seller is rewarded for closing the deal in front of them and will therefore build the tier that closes it, which is by construction the cheapest tier that includes everything. In a solo business all four roles are the same person, and the discipline that replaces the separation is the calendar: contents and prices are decided at the annual review, written down with a date, and not revisited during a negotiation. In an established company a short written procedure records who proposed, who costed and who approved, and Chapter 10 puts it on a cadence.

5.3 Good, Better, Best

5.3.1 Three is the working number, and why

The three-rung ladder is the most common price structure in business because it does three jobs at once. It gives a buyer who cannot afford the main offer somewhere to start rather than somewhere else to go. It gives a buyer who needs more a way to spend more without a negotiation. And it turns the buyer's question from "yes or no" into "which one." Book E1, Chapter 7, presents three options in the sales conversation and observes that the middle one is chosen most often; this chapter builds the options that conversation presents, which is a different and earlier piece of work. The ladder is no longer a software habit either: the second edition of The Price Advantage describes its own revision as extending to the pricing of tiered services, alongside software and configured products (Baker et al., 2010).

Three is a working number rather than a magic one, and the argument for stopping there is arithmetic rather than psychological. Each tier is a set of contents to maintain, a cost of service to measure, a fence to enforce, and a line in every quote and every support conversation. More consequentially, the number of pairs a seller must be able to distinguish grows faster than the number of tiers: three tiers give three ordered pairs to explain, four give six, and five give ten. Book B1 (Marketing Fundamentals for Business Owners), Chapter 6, states the same conclusion in one line — too many tiers paralyze — and this book offers no measured study of choice overload in support of it, because none was read for this volume. What it has is the pair arithmetic above and the observation that every tier a business cannot describe in one sentence is a tier it will end up discounting.

WATCH OUT Too many tiers do not segment a market; they paralyze it and then they leak The temptation when a ladder underperforms is to add a rung, and it is almost always the wrong move. Adding a fourth and a fifth tier does four things, of which only the first is desirable. It captures a little more from a narrow band of buyers between two existing rungs. It multiplies the pairs a seller must distinguish, from three at three tiers to ten at five. It adds fences, and every fence that is not enforced becomes a discount granted automatically to whoever notices it. And it moves the conversation from the offer to the comparison, so that the buyer spends the meeting choosing among your lines instead of deciding whether to buy at all. There is a reliable symptom: the seller starts saying "most people take the middle one" as the opening sentence rather than the closing one. The remedy is subtraction — withdraw the rung chosen least, publish the withdrawal with an effective date, keep existing customers on their line until renewal, and measure the average revenue per customer for two quarters before concluding anything. Chapter 8 treats the mechanics of withdrawing a tier that customers already hold, which is a price change and has to be run as one.

5.3.2 Setting the gaps, the contents and the margin of each rung

A ladder is built from the middle outward. The middle rung is the one most buyers should choose, so it is priced first, against the value estimate of Chapter 2 and the alternatives measured there. The bottom rung is then built downward by removing something real and the top rung upward by adding something real, with the gaps set by what was removed and added rather than by a ratio.

On the case the gaps are $40.00 from Essential to Plus and $50.00 from Plus to Portfolio, which is 44.94% and 38.76% in relative terms, and the top rung is a little more than twice the bottom one. Those proportions are a consequence of the design rather than a target. Two figures discipline every rung: the cost of service, $15.00, $31.00 and $38.00 a month, and the contribution, which is the price less that cost and comes to $74.00, $98.00 and $141.00.

The table that follows gives the three tiers with everything a pricing decision needs. Its columns give the tier, its monthly price for one building, the monthly cost of serving it, the contribution in dollars, the margin rate, and the share of new buildings the model assumes will choose it. Watch the margin rate, because it does something counterintuitive that Section 5.3.3 explains.

Table 5.3 — The three tiers of the case, priced, costed and weighted

5.3.3 Rank tiers by contribution in dollars, never by margin rate

Read the margin rate column of Table 5.3 downward and it falls, then rises: 83.15%, 75.97%, 78.77%. The middle tier, the one the business most wants buyers to choose, has the lowest margin rate of the three. That is not a mistake but a consequence of the cost of service, which more than doubles from $15.00 to $31.00 between Essential and Plus while the price rises by less than half.

A business that manages by margin rate will therefore push Essential, and it will be wrong every time. The correct ranking is by contribution in dollars per buyer, which rises across the ladder: $74.00, $98.00, $141.00. A building on Plus is worth $24.00 a month more than one on Essential, notwithstanding the margin-rate points it gives up, and a building on Portfolio is worth $67.00 more. Margin rate is a useful diagnostic of whether a tier is priced against its own cost; it is a destructive way to choose which tier to sell, because it steers the sales effort toward the smallest customers. Book G1 (Accounting Fundamentals for Business Owners) computes the margin; this book ranks the ladder by dollars.

Figure 5.1 shows the three tiers as three columns, each split into the monthly cost of service at the bottom and the monthly contribution above it, with the price marked at the top, the assumed share of new buildings written under each column, and the resulting average revenue per building on the right.

Figure 5.1 — Three tiers, what each one costs to serve, and what the mix averages. Each column is one tier of the case, measured in dollars a month for one building: Essential at $89.00, of which $15.00 is the monthly cost of service and $74.00 is contribution, a margin rate of 83.15%; Plus at $129.00, of which $31.00 is cost of service and $98.00 is contribution, a rate of 75.97%; and Portfolio at $179.00, of which $38.00 is cost of service and $141.00 is contribution, a rate of 78.77%. Under each column is the assumed share of new buildings, 55% for Essential, 35% for Plus and 10% for Portfolio. On the right, that mix gives an average revenue per building of $112.00 a month against an average cost of service of $22.90 and a contribution of $89.10, an uplift of 41.77% over the single earlier price of $79.00, of which $10.00 comes from the entry price and $23.00 from the mix. Hydrosentinelle inc. is fictional, and the prices, the service costs and the mix are illustrative assumptions of the model.

DO THE MATH What the tier mix is worth, and where the uplift comes from Fictional example; the prices, the service costs and the mix are illustrative assumptions of the model. Take the three published tiers at $89.00, $129.00 and $179.00 a building a month, and the assumed division of new buildings, 55%, 35% and 10%. The average revenue per building is $89.00 times 55%, plus $129.00 times 35%, plus $179.00 times 10%, which is $112.00 a month. The average cost of service is $15.00 times 55%, plus $31.00 times 35%, plus $38.00 times 10%, which is $22.90, so the average building contributes $89.10 a month. Against the single earlier price of $79.00 the average is up $33.00, an uplift of 41.77%. Now attribute it: had every new building simply moved to the new entry price of $89.00, the average would have been $89.00 and the gain $10.00, so the remaining $23.00 is the mix, which is to say the structure. Two cautions travel with these numbers. The mix of 55%, 35% and 10% is an assumption of the model rather than a measurement of the market, and Chapter 9 sets out how a business of this size measures the mix it gets. And the 41.77% applies to new buildings; the two hundred and six already on $79.00 are a separate operation, priced and sequenced in Chapter 8.

5.3.4 An option that nobody chooses is badly built, and there are four ways

Book E1, Chapter 14, sends the reader who has "options that nobody chooses" to this book, and Chapter 3 promised that this chapter would say which thing is wrong. There are four possibilities, they are distinguishable, and the diagnosis takes an afternoon.

• The fence is invisible. Buyers cannot tell what separates the unchosen rung from its neighbor, so they default to the cheaper one. The test is the read-aloud test of Section 5.2, and if the reader hesitates the remedy is to rewrite the contents and the sentence rather than the price.
• The gap is wrong. The fence is clear and the contents are real, but the extra money buys less than it costs the buyer to justify. Compute what the extra contents are worth using the economic-value method of Chapter 2 and compare that with the gap: if the gap exceeds the value the price is wrong, and if it does not, the problem is elsewhere.
• The segment does not exist. The rung was built for a kind of buyer the business does not actually meet. The test is to name three real prospects from the last year who would have bought it. If the list cannot be filled, the rung is a hypothesis that has been falsified, and it should be withdrawn.
• It is never presented. The rung exists on the price page and never reaches a quote, because the seller stopped offering it after the first refusal. Pull the last year of proposals and count how many contained the rung at all; Chapter 9 sets out the measurement, and it is the first to run because it is the cheapest and it disqualifies the other three.

Only the second of these four is repaired by changing a price. That is worth stating plainly, because the reflex when an option does not sell is to cut its price, and cutting the price of an option whose fence is invisible simply moves the leak one rung up the ladder.

METHOD Designing and pricing a three-tier ladder in eight steps When to use it: when a business with a single price is moving to a structure, and again at every annual price review, before any number is published. Steps: (1) Choose the price metric first, using the four tests of Section 5.4, and write it in one sentence, because every price below is expressed in that unit. (2) Obtain the cost of serving one unit at the level of service intended for the middle rung, from the accounting records rather than from memory. (3) Price the middle rung against the economic value estimate of Chapter 2 and the alternatives measured there, not against the cost, and write the sentences that justify it to a buyer. (4) Build the bottom rung by removing contents that are genuinely separable, cost the reduced service, and price it so that its contribution in dollars is positive. (5) Build the top rung by adding contents a named kind of buyer wants, cost the richer service, and set its price so that its contribution exceeds the middle rung's by more than the extra cost of serving it. (6) Write the fence between each adjacent pair as one sentence, say which of the six kinds it is, and say how it will be verified. (7) Set the minimum, and check that the smallest customer it admits still contributes more than it costs to serve. (8) Compute the weighted average revenue and cost per unit under an explicit mix assumption, and publish the list with an effective date. Deliverable: a one-page price list giving the metric, the three rungs with their contents, prices, costs of service and contributions, the two fences in one sentence each, the minimum, the mix assumption and the effective date. Verification criterion: a person who did not build the list can state, for each rung, who should buy it and what separates it from its neighbor, and the weighted average of step 8 reproduces exactly from the prices and the mix written on the page.

5.4 Per-Seat, Per-Unit, Per-Site — and the Minimum

5.4.1 The metric is the decision you live with longest

Of every choice in this chapter, the price metric is the hardest to change later and the one that most determines whether a business grows with its customers. A metric decides whose expansion the seller is paid for: software charged per seat grows when the customer hires, and software charged per company does not. Book B1, Chapter 6, describes per-unit and per-user pricing as the structure a buyer perceives as fair because it scales with their size, and adds the warning this section is built around: small customers may be unprofitable, so set a minimum.

Four tests decide a metric. Does it track the value the buyer receives, so that a buyer who gets more pays more without feeling punished? Can the buyer count and predict it, since a buyer who cannot forecast next year's bill will discount the whole offer mentally for the uncertainty? Can the seller measure and audit it without a dispute? And does it track the cost to serve, so that the customers consuming the most capacity are the ones paying the most?

The table that follows applies those tests to the three metrics Hydrosentinelle could plausibly have used. Its columns give the candidate metric, whether it tracks the value, whether the buyer can count it, whether it tracks the cost to serve, and what it would have done on the case. The last column holds the arithmetic that decided the question.

Table 5.4 — Three candidate price metrics for the case, tested

5.4.2 The minimum, and the arithmetic that sets it

The third row of Table 5.4 holds the general lesson in one number. Priced per apartment at the Essential rate, a six-apartment building would pay $11.10 a month against a cost of service of $15.00. That customer would be served at a loss forever, would be entirely satisfied, and would never complain, which is exactly why this kind of loss survives for years. A metric that can produce a bill below the cost of service needs a minimum, and the minimum is not a marketing device but an arithmetic floor.

A minimum takes one of three forms. A minimum quantity requires the buyer to purchase at least a stated number of units, which is cleanest when the units are what the buyer counts. A minimum charge accepts any quantity but bills at least a stated amount, and is the form most often used in professional services. A minimum commitment accepts any quantity at the list rate but requires a term, so that the fixed cost of setting the customer up is recovered over a known period. Hydrosentinelle uses the first form implicitly: the smallest purchase is one building, and one building at the entry tier contributes $74.00 a month against a cost of $15.00, so no minimum charge is needed on top.

The arithmetic that sets a minimum is always the same. Identify the part of the cost to serve that does not shrink when the customer is small — the onboarding, the account record, the monthly statement, the support availability. Divide it by the period over which you intend to recover it, and add the target contribution: that is the minimum charge. If the published metric can generate a bill below it, the minimum must appear beside the metric in the same type size, because a minimum discovered at the invoice stage is a surprise charge, and Chapter 7 explains why a surprise charge is a legal problem and not merely a commercial one.

CASE IN POINT Why the Portfolio tier requires five buildings Fictional example; names and figures are invented. The Portfolio tier of the case is available only to a buyer with at least 5 buildings under subscription, and the founders were asked twice in the first quarter to waive the condition for a single large building. They did not, and the reasoning is a pricing argument rather than a policy one. Portfolio exists because a buyer managing several buildings has a problem a single-building buyer does not: one view across the portfolio, one point of contact, and a response commitment that holds when three buildings alarm in the same week. That is what the extra $7.00 of monthly cost of service buys, taking the tier from $31.00 to $38.00 a building, and it is why the tier is priced at $179.00 rather than $129.00. A buyer with one building cannot use any of it, so selling Portfolio to that buyer would be selling $50.00 a month of contents the buyer cannot consume. The minimum of 5 buildings is also the threshold at which the volume table of Section 5.5 begins, so the sales conversation has one number in it rather than two. Laurence Bergeron, the president, insisted on a second property: the count is taken from Hydrosentinelle's own subscription records rather than from the buyer's declaration, so no one has to prove anything and no one can qualify by promising buildings that never arrive. The condition is printed above the price, in the same type size, and the two buyers refused in the first quarter both took Plus for their single building and were told in writing what would change at the fifth.

5.4.3 Per-seat pricing outside the case, and the hybrid it has become

Most readers will not be charging per building. The two metrics that dominate small-business software and services are the seat and the transaction, and each has a characteristic failure. Per-seat pricing fails when value is not proportional to headcount, which produces the pathology of shared logins: the buyer takes three seats and gives the password to nine people, and the seller's revenue stops tracking anything. Per-transaction pricing fails in the opposite direction, when usage is seasonal or volatile, because a bill that triples in a good month teaches the buyer to fear success.

The practical answer is a hybrid: a per-seat or per-site base that provides predictability, plus a usage component that captures the customers who consume far more than the average. Chapter 4 treats the usage component in full, including how the included allowance is set. The structural point that belongs here is that a hybrid needs two fences rather than one — the fence between tiers and the boundary of the included allowance — and each must be as clearly stated as the price itself.

WHAT CHANGED Subscription businesses stopped relying on a single billing model Ten years ago the structural question for a subscription business was answered once and then left alone: one metric, one billing period, one list. That has measurably changed. ChartMogul's billing report, published on May 18, 2025 and covering the full year 2024, analyzes more than 2,500 subscription software companies and excludes from parts of its analysis the companies that rely one hundred percent on a single billing model — the hybrids had become numerous enough that the single-model firms had to be separated out. The same report finds that the billing mix varies with the size of the customer rather than with the taste of the seller: among companies with average revenue per account of $1,000 or more, 55% of revenue came from annual plans against 34% from monthly, while among companies with $15 million to $30 million of annual recurring revenue the monthly share dominated at 72% against 28%. Two cautions belong with these figures: the data are aggregated billing records from one vendor's own customers, a non-probability sample rather than a survey of the industry, and the report excludes companies below $300,000 of annual recurring revenue, so it says nothing directly about the smallest businesses. What a small business should take from it is not a benchmark but a permission: running more than one structure at once is now ordinary practice, provided each structure has its own fence and its own arithmetic (Faustino, 2025).

5.5 Volume and Tiered Discounts

5.5.1 A volume discount is a structure, not a concession

A volume discount published in a table is a different object from a discount granted in a negotiation. A published table is part of the structure: it applies to everyone who meets the stated quantity, it can be verified, it is consistent between sellers, and it can be defended as the ordinary terms of the business. A discount granted in a conversation is policy, it belongs to Chapter 6, and the case is used there to show that six such lines removed 30.59% of the list margin of a single deal.

A volume table is also a price in its own right, and it has to be justified. The justification has two legitimate sources. The first is cost: a buyer who takes twenty kits on one order costs less per kit to sell to, to schedule, to ship and to install than twenty buyers taking one each. The second is commitment: the buyer is promising volume, and the promise is worth something. What is not a justification is the size of the buyer as such. A table that gives a large buyer a better price simply because the buyer is large, with no cost difference and no commitment, is a discount with a table around it, and in the United States it can be worse than that.

Hydrosentinelle's table has three bands. Buyers taking 1 to 4 kits on one order pay the list price of $4,800.00. Buyers taking 5 to 19 kits pay 6% less, which is $4,512.00. Buyers taking 20 or more pay 12% less, which is $4,224.00. The cost of an installed kit is $2,100.00 and does not change with the order size in the company's current operations, which means the whole of the discount comes out of contribution.

The table that follows sets out the three bands with what each one leaves. Its columns give the band, the kit price, the discount, the contribution per kit, and the additional volume the band must generate to be worth as much as the band above it in total contribution. The last column is the question a volume table must answer and almost never does.

Table 5.5 — The volume table of the case, and what each band must buy back

The last column is computed the way Chapter 2 computes every break-even volume change: the undiscounted contribution of $2,700.00 divided by the discounted contribution. For the 6% band, $2,700.00 divided by $2,412.00 is 1.1194, so the band must produce 11.94% more kits merely to stand still; for the 12% band, $2,700.00 divided by $2,124.00 is 1.2712, so it must produce 27.12% more. A table published without asking those questions is a hope with percentages in it.

5.5.2 The cliff hiding in an all-units table

Hydrosentinelle's table, like most volume tables written quickly, is an all-units table: crossing a threshold changes the price of every unit on the order, not merely the units above the threshold. That is simple to explain and it contains a trap that the arithmetic exposes immediately.

Consider a buyer with nineteen buildings and a buyer with twenty. Nineteen kits at $4,512.00 come to $85,728.00; twenty kits at $4,224.00 come to $84,480.00. The twentieth kit does not merely earn nothing, it reduces the order by $1,248.00, and on contribution the damage is larger because that kit also carries $2,100.00 of cost: nineteen kits contribute $45,828.00 and twenty contribute $42,480.00, so selling one more kit costs the business $3,348.00.

The table that follows sets the two orders side by side. Its columns give the measure, the nineteen-kit order, the twenty-kit order and the difference. Everything in it follows from the published table, which is the point: nothing was negotiated and nobody made a mistake.

Table 5.6 — What a nineteen-kit order and a twenty-kit order are worth under an all-units table

There are three remedies, and a business should choose one deliberately rather than discover the cliff in an invoice. An incremental table applies the discount only to the units above each threshold — the first four kits at $4,800.00, the next fifteen at $4,512.00, everything beyond at $4,224.00 — so a twentieth kit always adds revenue and the cliff disappears by construction. A better-of rule quotes whichever of the two calculations is better for the buyer, which keeps the simplicity of an all-units table and removes the perverse zone below each threshold. Moving the thresholds to quantities the business never sees is the third, and it is only honest if that is really true. Hydrosentinelle publishes an all-units table and applies the better-of rule in its quote template, and Chapter 7 shows the twelve-building quotation that results.

THE RULE Robinson-Patman Act, 15 U.S.C. § 13: the same goods at different prices to competing buyers In the United States, the Robinson-Patman Act makes it unlawful to discriminate in price between different purchasers of commodities of like grade and quality where the effect may be substantially to lessen competition or tend to create a monopoly in any line of commerce. Three features of the text govern how a volume table should be built for American buyers. First, the provision reaches tangible goods, not services, so a services business is outside it while a business shipping equipment across the border is not. Second, the statute contains a cost-justification defense: a price difference is permitted where it makes only due allowance for differences in the cost of manufacture, sale or delivery resulting from the differing methods or quantities in which the goods are sold or delivered, which is precisely a volume discount grounded in a measured cost difference, and the demonstration required is heavy; a second defense covers price changes made in response to changing conditions affecting the market for or the marketability of the goods, such as perishable or seasonal goods. Third, once a discrimination is shown the burden of rebutting the case rests on the person charged, and the recognized answer is that the lower price was made in good faith to meet an equally low price of a competitor — to meet it, not to beat it. The Act also reaches promotional allowances and services, which must be available on proportionally equal terms to competing customers, and it reaches the buyer, making it unlawful knowingly to induce or receive a discriminatory price. Scope and date: United States commerce; text read on September 19, 2026 at the Legal Information Institute of Cornell Law School, the official electronic code having returned an anti-robot redirection that was not circumvented. Verify the current text before relying on it.

5.5.3 Cumulative volume, committed volume and the promise that is not a quantity

A volume table can count in three ways, and the difference decides who carries the risk. An order-based table counts the units on a single order, which is simplest to administer and is what the case uses. A cumulative table counts units bought over a period, usually a year, and rewards a customer for staying; the seller must decide in advance what happens when the year ends short of the threshold, and the honest answers are a retroactive adjustment or nothing at all, never a quiet continuation of the discount. A committed table gives the discount up front against a contractual promise of volume, and it is the only one of the three in which the seller is exposed.

The rule for a committed table is short: if the commitment is not enforceable, it is not a commitment and the discount is a gift. Enforceable means that the contract states what happens if the volume does not arrive — a retroactive recovery, a shortfall payment, a return to list price — and that the business is willing to invoke it. Book H1 (Contracts and Commercial Agreements) drafts that clause; a business that has never enforced one should price as though the commitment did not exist.

QUÉBEC & CANADA NOTE Volume tables, suggested resale prices and what the Competition Act reaches Nothing in the federal provisions read for this book requires a Canadian seller to charge every buyer the same price, and a published volume table is an ordinary commercial practice. Two provisions bear on it. The first is Competition Act, section 45, which makes it a criminal offense to agree with a competitor to fix, maintain, increase or control prices; a volume table coordinated among competing suppliers is not a structure but a conspiracy, and Chapter 6 treats that provision in full. The second is section 76, price maintenance, which reaches a supplier who, by agreement, threat, promise or any like means, has influenced upward or discouraged the reduction of the price at which a customer or reseller supplies, offers or advertises a product in Canada, or who has refused to supply a person because of that person's low pricing policy — but only where the conduct has had, is having or is likely to have an adverse effect on competition in a market. Both limbs are required. Section 76 also settles the question every supplier with resellers asks: suggesting a resale price is evidence of upward influence unless the supplier makes it clear that the recipient is under no obligation to accept it and that business relations will not suffer in any way, and advertising a resale price is evidence of influence unless the advertisement makes it clear that the product may be sold for less. The Competition Bureau's enforcement guidelines add that the adverse-effect threshold is lower than the substantial-lessening test used elsewhere in the Act, that market power is required for an adverse effect, and that the Bureau will not generally examine market power below a 35% market share — a statement of enforcement policy, not an immunity, and not a threshold found in the Act (Competition Bureau Canada, 2014). In Québec, a volume table offered to businesses, as in the case, sits outside the consumer rules; when the same list is shown to consumers, the advertised-price obligations of Chapters 3 and 7 apply on top.

5.6 Add-Ons, Options and the Attach Rate

5.6.1 What an add-on is for

An add-on is a separately priced item a buyer may purchase alongside the main offer and would have no reason to buy on its own. It lets the main offer stay simple, because the needs of the minority are not built into the price everyone pays; it captures value from buyers whose situation is larger than average without raising the entry price; and it gives the seller something to trade in a negotiation that is not a discount, a point Chapter 6 returns to. It also carries a danger: if the contents that made the middle rung worth choosing are gradually pulled out and sold as add-ons, the ladder collapses into an entry price plus a shopping list. The discipline is one sentence: an add-on extends the offer for a minority, and it never removes something the tier promised.

5.6.2 The sensor pack, priced and measured

Hydrosentinelle's add-on is a sensor pack. The installed kit contains 24 sensors, which covers the great majority of buildings the company serves. A building with more risk points — additional mechanical rooms, a second basement, a rooftop unit — needs more, and the pack adds 6 sensors for $390.00 against a cost of $120.00, leaving a contribution of $270.00, which is 69.23% of the pack's price.

Two comparisons are worth making, because both are objections a buyer will raise. The first is per-sensor: the kit works out to $200.00 a sensor and the pack to $65.00, which looks like an inconsistency until one remembers that the kit also carries the gateway and the installation visit, costed in Book A5 at $1,650.00 of hardware and $450.00 of installation within the $2,100.00 total. The answer is a real one and should be written on the price page rather than improvised. The second comparison is with the subscription: the pack is a one-time charge and does not change the monthly tier, because the tier is priced per building and the metric is not the sensor. A business that lets an add-on quietly change the metric has two metrics and no structure.

The measurement that governs an add-on is the attach rate: the share of primary sales that carry it. Every pack attached to a kit adds $270.00 of contribution, so a pack on one kit in ten adds $270.00 for every ten kits sold, while a pack on every kit adds $2,700.00 for every ten, which is exactly the list contribution of one additional kit. Put on the year, the case plans 180 kit installations, so if every one carried a pack the packs would add $48,600.00 of contribution, 6.92% of the year's gross margin of $701,808.00, without one new customer being found. That is an upper bound and is stated as one: the company has not yet measured its attach rate, and Chapter 9 sets out the measurement, which requires nothing more than counting the packs sold and dividing by the kits installed.

5.6.3 Where an option belongs on the page

Three rules of placement follow. An add-on is shown after the tier it extends, never beside it, because a buyer still choosing a tier should not be choosing among five things. Every add-on carries its own cost in the internal price list even when the buyer never sees it, because an add-on whose cost is unknown is the easiest thing in a business to give away. And an option chosen by nearly every buyer is not an option: it is part of the offer left outside the price, and the honest move is to fold it into the tier and raise the tier, which is a price change and belongs to Chapter 8.

TOOLS OF THE TRADE What a catalog must hold before a tiered offer can be quoted twice the same way Whatever software a business uses, a tiered structure needs a catalog that holds a specific set of things, and the absence of any one of them shows up later as a price that two people quote differently. It must name the price metric once, in a field of its own, so that every line is expressed in the same unit. It must hold one record per tier, carrying its contents, its price, its cost of service, its fence written as a sentence and an effective date, with previous versions retained rather than overwritten, because a price without a history cannot support any claim about what the ordinary price was. It must hold a volume table with its band boundaries and an explicit flag saying whether the bands are all-units or incremental, since the two produce different invoices from the same numbers. It must hold add-ons with their own prices and costs, linked to the tiers they extend, the minimum in the same place as the metric, and the currency of each price. It must produce a dated price list as a document, because that is what a seller quotes from and what an auditor asks for. Finally, it must report the mix: how many customers sit on each tier and what the average revenue per unit is, which is the measurement Chapter 9 uses. A spreadsheet with an effective-date column and a change log satisfies every requirement here; the point is the fields, not the software.

5.7 Hourly, Daily and Fixed-Price Work

5.7.1 What an hourly rate really prices

An hourly or daily rate prices effort, and it is the structure a business falls back on when the scope cannot be described in advance. Its virtue is that the seller cannot lose money on a job. Its costs are easy to underestimate: it caps income at the hours available, so the business grows only by adding people or raising the rate; it invites scrutiny of every hour, converting the relationship from an outcome into a timesheet; and it penalizes efficiency, since the more competent the seller becomes the fewer hours the work takes. Book B1, Chapter 6, states the same three points and adds the remedy: state an estimate and a cap.

The estimate and the cap together are what make an hourly arrangement buyable. An estimate without a cap transfers all the risk of the unknown to the buyer, who will therefore either refuse or build a private discount into the decision. A cap without an estimate tells the buyer nothing about the likely bill. Stated together — this work is estimated at a stated number of hours and will not be billed above a stated ceiling without written agreement — they give the buyer a worst case and the seller the right to stop and renegotiate rather than absorb an overrun silently. The clause that expresses it belongs to Book H1; the decision to offer it belongs here.

5.7.2 A fixed price is an hourly price with the risk priced in

A fixed price per project or per package prices an outcome, and it transfers the risk of overrun from the buyer to the seller. That transfer is the product: a buyer who accepts a fixed price is buying certainty, which is why a correctly priced fixed-price offer is almost always more expensive than the expected hourly bill for the same work. The gap is not greed; it is the price of the risk.

That price can be computed rather than guessed, and it needs two numbers from the business's own records: how often the work runs over, and by how much when it does. Multiply the two and add the result to the estimate. A business that cannot supply them should quote hourly with a cap until it can, because a fixed price set without them is a wager. On the case the risk is small, real and documented: installations are budgeted at $450.00 within the kit cost of $2,100.00, and the deal examined in Chapter 6 carried an installation overrun of $150.00 that was never billed. That one unbilled overrun is a third of the budgeted installation cost given away, and a business that absorbs it on a predictable share of jobs without ever counting the share has an unpriced risk sitting inside a fixed price.

The table that follows compares the three arrangements on the dimension that actually distinguishes them, which is who carries what. Its columns give the arrangement, who carries the risk that the work runs long, what the buyer has to accept, what the seller must be able to produce, and the situation each one fits. Read the fourth column before choosing, because it is the column that says whether the business is ready for the arrangement at all.

Table 5.7 — Hourly, capped and fixed: who carries the risk of a scope that drifts

5.7.3 Scope drift, and the change process that prices it

The failure mode of a fixed price is not the price; it is the scope. Work expands because buyers discover things, because the seller wants to be helpful, and because the boundary was written loosely enough that both parties can read it their own way. Three habits contain it, and none of them is contractual in the first instance.

• The scope is written as what is included and what is excluded, in that order, with the exclusions as specific as the inclusions. A scope that lists only inclusions is read by the buyer as including everything not explicitly refused.
• Every change is priced before it is done, in writing, even when the change is small and even when the relationship is good. A business that prices changes after the fact will not price most of them at all, and the sum of the ones it did not price is the margin it wonders where it went.
• The change price uses the same structure as the original. A fixed-price project whose changes are billed hourly has taught the buyer that the cheap way to buy is to order a small project and then change it, which is precisely the behavior a fixed price was meant to prevent.

On the case, this is why the installed kit is sold at a fixed price with installation included, while electrical work outside the mechanical room — which cannot be scoped from a photograph and varies by an order of magnitude between buildings — is quoted separately by a licensed contractor, with an estimate and a cap. That is the structure rather than a compromise of it: a fixed price for what is repeatable, capped hourly for what genuinely is not.

5.8 Assembling a Price Architecture

5.8.1 The seven structures, and where this book carries each

Book B1, Chapter 6, compares seven pricing structures a small business chooses among and promises that this book will supply the methods behind the numbers of each. The promise is discharged across two chapters. Chapter 4 carries the structures that are revenue models: the recurring subscription and the retainer, usage-based charging, performance-based pricing combined with a base fee, and freemium with its cost of serving free users and its measured conversion. This chapter carries the structures that are shapes of a price: the hourly and daily rate with its estimate and cap, the fixed price per project or package with the risk of scope drift priced in, the good-better-best ladder, per-seat and per-site charging with its minimum, and the volume table.

The table that follows gives all seven with the chapter that carries each. Its columns give the structure, what it charges for, where it is treated here, and what this book adds. The last column is the reason the reference exists: Book B1 named the structures and their signals, and this book supplies the arithmetic underneath them.

Table 5.8 — The seven structures of Book B1, Chapter 6, and where this book carries each

5.8.2 The combination, computed on the case

Book B1, Chapter 6, concludes that most small businesses end up with a combination: a fixed-price entry package, a retainer or subscription for the continuing relationship, and hourly or per-unit pricing only for work that genuinely cannot be scoped. The case is exactly that combination, and it can be put in numbers for one new building in the third year.

Under the old architecture a new building bought an installed kit at $4,800.00 and a subscription at $79.00 a month, so the first year was $5,748.00 of revenue and $3,468.00 of contribution, being $2,700.00 on the kit plus twelve months at $64.00, a first-year contribution rate of 60.33%. Under the new architecture the same building buys the same kit and an average subscription of $112.00 a month, so the first year is $6,144.00 of revenue and $3,769.20 of contribution, being $2,700.00 plus twelve months at $89.10, a rate of 61.35%. The combination adds $396.00 of first-year revenue and $301.20 of first-year contribution per building, through the shape of the list rather than any change to the kit, the cost or the sales effort. The third element of B1's combination, the hourly work, contributes nothing to that arithmetic and is not meant to: out-of-scope electrical work is quoted separately with a cap precisely so that it never distorts the two prices that matter.

5.8.3 What a price architecture looks like when it is finished

A finished architecture is one page holding nine things: the price metric, stated once; the tiers with their contents, prices and costs of service; the fence between each adjacent pair, in a sentence; the volume table with its bands and an explicit statement of whether they are all-units or incremental; the add-ons with their prices and costs; the minimum; the currency; the effective date; and the name of the person who approved it. Everything else belongs elsewhere. Discounts, authority and the floor are policy and live in the one-page discount policy of Chapter 6; how the list is displayed is communication and lives in Chapter 7; when the list changes and how customers are moved is a price change and lives in Chapter 8. Keeping those three out is what keeps the architecture on one page.

Two habits keep it true after publication. The first is the effective date, written on the page and never overwritten, with superseded versions kept, because a business that cannot show what its price was on a given date cannot support a claim about a previous price or settle a dispute about what a customer signed. The second is the annual review, at which the metric is questioned, the costs of service are refreshed from the accounting records, the mix is measured against the assumption, and the fences are tested by reading them aloud to someone outside the business. Chapter 10 puts that review on a cadence with an owner and a definition of done, and Chapter 9 supplies the measurements it needs.

IN LA MACHINE One price and one cost per package, and no tier field Read in the application's source at its commit of September 7, 2026, a session package (Business → Session Packages, business/packages) carries a name, a type of meetings or minutes, a quantity, an unlimited flag, one price, one cost and a currency, and a subscription package (business/subscription-packages) carries a name, one price, one cost, a description, an image, an active flag and a billing period that is monthly or annual. Both compute the profit per sale and the margin percentage on screen, which is genuinely useful and is more than most small businesses do on paper. What the code does not carry is the rest of a structure. There is no tier field, so the three rungs of a ladder can only be entered as three unrelated records with nothing linking them, ordering them or holding the fence between them; there is no price metric field, no minimum, no volume band for the reader's own offer, no add-on relationship and no attach-rate report; and there is no effective date and no price history, so changing a price overwrites the previous one. The only quantity breaks in the platform live under business/partner-programs, and they price La Machine's own distribution rather than the reader's offer. The practical consequence for a reader following this chapter is that the one-page architecture of Section 5.8.3 has to live outside the platform, while the price and cost of each rung can usefully live inside it (La Machine, 2026).

Costly Mistakes

• Treating a new price list as a price increase, or a price increase as a new list. These are two decisions with two different risks, and on the case $10.00 of the $33.00 of uplift came from the level while $23.00 came from the structure; a business that cannot say which is which can neither repeat its success nor diagnose its failure.
• Publishing a second price without writing the fence first. A tier that a buyer cannot tell apart from the next one is not a segment of the market but a discount granted to everyone who reads carefully, and the average price of the business falls to its lowest published line.
• Ranking tiers by margin rate instead of contribution in dollars. On the case the middle tier has the lowest rate at 75.97% and the second-highest contribution at $98.00 a month, so a business managed by rate would push the tier worth $74.00 and call it discipline.
• Publishing an all-units volume table without testing the thresholds. Under the case's own table, a twentieth kit reduces the order by $1,248.00 and the contribution by $3,348.00, and nobody negotiated anything to make that happen.
• Choosing a price metric that can bill below the cost of service. Priced per apartment at the Essential rate, a six-unit building would pay $11.10 a month against a cost of $15.00, and the customer would never complain about it.
• Quoting a fixed price without a measured overrun. A fixed price is an estimate plus the price of a risk, and a business that has not counted how often its work runs long and by how much is not pricing that risk but betting on it.

Key Takeaways

• Structure and level are two decisions, and the structure is usually the larger one. Moving the case from one price to three raised the average revenue per building from $79.00 to $112.00, an uplift of 41.77%, of which $23.00 of the $33.00 came from the shape of the list.
• A fence is what makes two prices legitimate rather than one price and a leak, and a usable fence is stated before the price, verifiable by the seller, and acceptable to the buyer when read aloud.
• A ladder is built from the middle outward, with the gaps set by what each rung contains and costs rather than by a ratio, and the case's gaps of $40.00 and $50.00 follow from contents and costs of $15.00, $31.00 and $38.00 a month.
• Rank tiers by contribution in dollars, never by margin rate, because the case's ladder contributes $74.00, $98.00 and $141.00 a month while its margin rate falls and then rises.
• An option that nobody chooses is badly built in one of four ways — an invisible fence, a wrong gap, a segment that does not exist, or a rung that is never presented — and only the second is repaired by changing a price.
• The price metric is the structural choice you live with longest, and a metric that can generate a bill below the cost of service needs a published minimum beside it.
• A volume band must buy back what it gives away, and the case's 6% and 12% bands need 11.94% and 27.12% more kits simply to leave the business where it was; a finished architecture states those bands, the metric, the tiers and their costs, the fences, the add-ons, the minimum, the currency and the effective date on one page.

Your Action Plan

1. Name your price metric in one sentence. The deliverable is a single written sentence saying what your price is charged per and how that unit is counted; it takes about twenty minutes, and it is done when two people in the business write the same sentence independently.
2. Test the metric against the four criteria. The deliverable is a four-line note saying whether your metric tracks value, can be counted and predicted by the buyer, can be measured by you without dispute, and tracks your cost to serve; it takes about an hour, and it is done when each line ends in a yes, a no, or a named remedy.
3. Compute the smallest bill your metric can generate, and set a minimum. The deliverable is two numbers, the smallest possible bill and the cost of serving the customer who would produce it, followed by the minimum you will publish; it takes about an hour with your cost records, and it is done when the minimum appears beside the metric on your price page in the same type size.
4. Write the fence between every adjacent pair of prices you publish. The deliverable is one sentence per pair, each naming the kind of fence and how it will be verified; it takes an afternoon, and it is done when someone outside the business reads the sentences without the prices and correctly says who should buy each line.
5. Cost each tier and rank them by contribution in dollars. The deliverable is a table like Table 5.3 giving each tier's price, cost of service, contribution and margin rate, plus the weighted average under a stated mix assumption; it takes about half a day with your accountant, and it is done when the weighted average reproduces exactly from the prices and the mix on the page.
6. Stress-test your volume table for cliffs. The deliverable is a two-column comparison of the order value and the contribution at one unit below and one unit above each threshold, with the remedy chosen and written down; it takes about an hour, and it is done when no quantity increase reduces either number, or the better-of rule is written into your quote template.
7. Assemble the one-page architecture and date it. The deliverable is a single page carrying the metric, the tiers with contents, prices and costs, the fences, the volume table with its all-units or incremental flag, the add-ons, the minimum, the currency, the effective date and the approver's name; it takes a day, and it is done when the previous version is archived rather than overwritten and the review date is in the calendar with an owner.

In La Machine

Every route named below was verified in the route inventory on September 19, 2026 and read in the product's source code at commit be8bf04 of September 7, 2026 (La Machine, 2026); nothing here is inferred from a page title.

• Business → Session Packages (business/packages, also reachable at productivity/packages) holds a package with a name, a type of either meetings or minutes, a quantity, an unlimited option, a price, a cost and a currency, and computes "Profit per sale" and "% margin" on screen; the quantity and the unlimited option are the only expression of structure the platform offers, and they are a useful one for a business selling blocks of time.
• business/subscription-packages holds a subscription package with a name, a price, a cost, a description, an image, an active flag and a billing period that is monthly or annual, which is enough to record a ladder as several separate packages and to see each one's margin.
• Business → Store → Products (business/shop and business/shop-product-edit) holds a product with a price, a currency, a status, a product cost, a shipping cost and shipping fees, and shows a profit panel computed from them, which is where an add-on such as the case's sensor pack would be recorded with its cost.
• Account → Profile → Revenues & Expenses (business/subscriptions) reads subscription transactions, package purchases and product purchases and writes each sale as a revenue line with its cost lines beneath it; it is the only view that places a revenue beside its margin, and therefore where the contribution ranking of Section 5.3.3 would be read.
• Control center → Projects (chains), Control center → Plan tasks and Control center → Today hold the annual architecture review of step 7 of the action plan as scheduled work with an owner and a due date, and Control center → Audit log records who changed what and when.
• What is not here. There is no tier field on a package or a subscription package: a ladder can only be recorded as several unrelated records, with nothing linking them, ordering them or holding the fence between them. There is no price metric field, no minimum, no product variant, no add-on relationship and no attach-rate measurement. There is no price list and no price book, no effective date on a price and no price history, so a superseded price is simply overwritten. There is no volume band and no quantity break for the reader's own products; the only volume tiers in the platform belong to La Machine's own distributor program under business/partner-programs, which sells La Machine rather than the reader's offer. There is no trial, no proration and no upgrade path between packages, and no link between a subscription package and any customer's actual subscription. There is no quotation of any kind — the page at business/quotes is a page of motivational quotes — and no invoice for the reader's own customers. Checkout is a demonstration (/checkout-demo) and /client-subscribe is disabled. The platform checks no price presentation against the Competition Act or the Consumer Protection Act, which Book B1, Chapter 6, states already and this book confirms. Book N8 will be the page-by-page guide to these screens; this section describes only what bears on a price structure.

Self-Check

1. What is the difference between a price structure and a price level, and how would you measure which of the two produced a change in your average revenue per customer?
2. What three properties make a price fence usable, and what happens to a published list when a fence has none of them?
3. Why should tiers be ranked by contribution in dollars rather than by margin rate, and what do the case's three tiers show?
4. What is the cliff in an all-units volume table, what does it cost on the case, and what are the three remedies?
5. An option on your price page has not been chosen in a year. What are the four possible causes, and in what order do you test them?

Answers. (1) The level answers "how much" while the structure answers what the price is charged per, how many prices there are, what separates them and what changes them automatically; measure the split by recomputing the average revenue per customer under the new structure with the entry price held at its old value, which on the case gives $10.00 of the $33.00 to the level and $23.00 to the structure (Sections 5.1.1 and 5.1.3). (2) A fence must be stated before the price, verifiable without an argument, and acceptable when read aloud; a list whose fences have none of those properties collapses to its lowest published price (Sections 5.2.1 and 5.2.3). (3) Because margin rate is driven by the cost of service and can fall on exactly the tier the business most wants sold: the rates run 83.15%, 75.97% and 78.77% while the contributions run $74.00, $98.00 and $141.00 (Section 5.3.3 and Table 5.3). (4) Under an all-units table the discount applies to every unit once a threshold is crossed, so an order can shrink as it grows: nineteen kits are worth $85,728.00 and twenty $84,480.00, a loss of $1,248.00 of revenue and $3,348.00 of contribution; the remedies are an incremental table, a better-of rule, or thresholds moved to quantities the business never sees (Section 5.5.2 and Table 5.6). (5) The fence is invisible, the gap is wrong, the segment does not exist, or the option is never presented; test the fourth first by counting how many quotes in the last year contained the option at all (Section 5.3.4).

Chapter References

Baker, W. L., Marn, M. V., & Zawada, C. C. (2010). The price advantage (2nd ed.). Wiley. https://www.wiley.com/en-us/The+Price+Advantage,+2nd+Edition-p-9781118267653 (publisher page read September 19, 2026; the second edition is by Baker, Marn and Zawada, Baker having replaced Roegner of the first edition)

Competition Bureau Canada. (2014, September 15; modified 2024, June 27). Price maintenance (section 76 of the Competition Act): Enforcement guidelines. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/en/how-we-foster-competition/education-and-outreach/price-maintenance-section-76-competition-act

Competition Act, RSC 1985, c C-34, ss. 45 and 76 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Faustino, S. (2025, May 18). SaaS billing report: Billing is the invisible force behind your growth. ChartMogul. Retrieved September 19, 2026, from https://chartmogul.com/reports/saas-billing-report/ (aggregated billing data from more than 2,500 subscription software companies for full-year 2024; a non-probability sample of one vendor's own customers, excluding companies below $300,000 of annual recurring revenue)

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Robinson-Patman Act, 15 U.S.C. § 13 (2024) (United States). Canonical source https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title15-section13; text read September 19, 2026, at the Legal Information Institute, Cornell Law School.

Chapter 6 — Discounts, Promotions and Negotiated Prices

The order that Karim Haddad closed in the second week of March was, by every measure the company used that day, a good one. A property manager with five buildings had signed for five kits of connected sensors at $4,512 each instead of the list price of $4,800, which is the published volume step for an order of five to nineteen buildings, and Hydrosentinelle inc. had beaten a competitor that had quoted lower. The invoice said six percent. Three months later Laurence Bergeron, preparing the annual review, pulled the same order apart line by line and found five other reductions that no one had called a discount: a further $225.60 conceded on the phone to end the negotiation, sixty days to pay instead of thirty, one free pack of six spare sensors, and an installation that ran long and was never billed. The price the company actually kept was $3,974.12, which is 82.79 percent of the list price, and the margin on that order was $1,874.12 instead of the $2,700 the price list promised. 30.59 percent of the margin had left the building, and the only line anyone could have found in the accounting system was the six percent. This chapter is about that gap. It measures what a discount costs before anything else is said about it; it builds the pocket price waterfall that makes the invisible reductions visible; it sets a margin floor and the authority to approach it; it states the rule that every discount is traded and never given; it designs, schedules and measures promotions; and it sets out the law that governs a promotional price and the two agreements a business may never make. Book E1 (The Sales Process: From Prospect to Signed Contract), Chapters 7 and 8, presents and defends a price in front of a buyer; this chapter builds the floor, the authority and the trades that the conversation is allowed to use.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Compute what any discount costs you, in dollars of margin and in the extra volume you would have to sell to replace it, using nothing but your price and your unit cost.
• Build a pocket price waterfall for one of your own transactions, so that every reduction between the list price and the money you keep is named, measured and attributed.
• Write a margin floor expressed as a margin rather than as a discount percentage, and measure it against the pocket price rather than against the invoice.
• Set discount authority by role, so that each person who negotiates knows exactly how far they may go, what they must record, and whom they must ask beyond that point.
• Trade every concession for something by working from a written list of counterparts, each one priced, rather than from whatever comes to mind under pressure.
• Run a promotion that is legal, dated and measured, and say before it starts how many extra units it must sell before it has paid for itself.

Why It Matters

Discounting is the only pricing decision most businesses make every week, and it is the only one most of them never write down. A price level is chosen once and reviewed rarely; a discount is granted on a Tuesday afternoon by whoever happens to be on the telephone, and it leaves no record that anybody ever reads. The arithmetic is unforgiving because a discount comes entirely out of margin and never out of cost. On the case in this book, a kit sold at list carries $2,700 of margin on a $4,800 price; a ten percent discount removes $480 of that margin, which is 17.78 percent of it, and replacing it would take 21.62 percent more units sold, as Chapter 2 computes. Nobody sells twenty-two percent more units to make up a ten percent discount. The discount is simply a transfer from the seller to the buyer, and the only question worth asking is what the seller received in exchange.

For a solo owner, the stakes are that the discount and the income are the same money. There is no one to hold the line, no approval step, and no colleague to say no on the owner's behalf; the floor has to be written in advance and treated as binding, because a number decided in a quiet moment is the only defense against a number requested in a loud one. For a small team, the stakes are consistency and knowledge. The moment two people may adjust a price, the business either has a policy or has a lottery, and the lottery produces a spread of prices for the same work that no one can explain to a customer who compares notes with another customer. For an established business, the stakes are that a discount is permanent. A reduction granted once becomes the reference price of that account for the life of the relationship, the account tells other accounts, and a business that has discounted systematically for three years cannot raise its prices; it can only stop discounting, slowly, deal by deal, against its own precedent.

There is also a legal dimension here that is sharper than anywhere else in this book. A promotional price is a representation to the public, and a comparison to a higher price is a representation about that higher price. In Canada the Competition Act treats a false or misleading representation as reviewable conduct, and as a criminal offense when it is made knowingly or recklessly; it also treats an agreement with a competitor about prices — including an agreement about discounts — as a criminal offense with no need to prove any effect on competition. A business can lose more on one badly documented crossed-out price than it earned on the promotion that carried it.

In This Chapter

• 6.1 What a Discount Actually Costs
• 6.2 The Pocket Price Waterfall
• 6.3 The Margin Floor
• 6.4 Discount Authority and Approval
• 6.5 The Rule of Exchange: Never Give, Always Trade
• 6.6 Promotions: Design, Calendar and Measurement
• 6.7 The Law of Promotional Prices
• 6.8 What You May Never Agree with a Competitor or Impose on a Reseller

6.1 What a Discount Actually Costs

6.1.1 A discount is not a percentage of the price; it is a percentage of the margin

The reason discounting feels harmless is that it is quoted in the wrong unit. A seller who offers ten percent hears a small number, because ten percent of a price sounds like a tenth of something the business has plenty of. The money does not come out of the price, however. It comes out of the margin, and the margin is a much smaller pile.

The arithmetic has two steps and one division. First, the dollars: a discount of ten percent on a kit priced at $4,800 is $480. Second, the comparison: the margin on that kit at list price is $2,700, because the unit cost is $2,100. The discount is therefore 17.78 percent of the margin, not ten percent of anything the business owns. The division that follows is the one that ends most arguments. To earn the same total margin after the discount, the business must sell enough additional units to replace $480 at the new unit margin of $2,220, which requires 21.62 percent more volume — the figure Chapter 2 computes as the break-even volume change of a ten percent discount. A sales team that could produce twenty-two percent more volume on demand would not need the discount.

The same arithmetic runs the other way for the subscription, and the contrast is instructive. The monitoring subscription is priced at $79 a month against a cost to serve of $15, so the margin rate is 81.01 percent. A ten percent discount there costs $7.90 of a $64.00 margin, which is 12.34 percent of it, and breaking even takes 14.08 percent more subscriptions. The general result is worth holding on to: the higher the margin rate, the less a discount costs in relative terms and the more volume the business would nonetheless have to find. A low-margin offer must almost never be discounted, because the volume required to repair the damage does not exist.

6.1.2 Four prices, and only one of them is real

Before anything can be measured, four numbers have to be given separate names, because businesses routinely use one word for all of them. The list price is the published price of the offer. The invoice price is what appears on the document the customer receives, after the discounts that the billing system knows about. The pocket price is what the business actually keeps once every other concession has been subtracted, including the ones that never touch an invoice. And the pocket margin is the pocket price minus the unit cost, which is the only margin figure that corresponds to money.

DEFINITION List price, invoice price, pocket price and pocket margin The list price is the published price of an offer before any reduction, and it is the reference against which every later reduction is measured; a business whose list price no longer corresponds to any actual transaction has lost its own reference point. The invoice price is the amount printed on the customer's invoice after the reductions the billing system records, which typically means published volume steps, promotional percentages and negotiated discounts entered as such. The pocket price is what the business actually keeps from the transaction after every remaining concession has been deducted, whether or not it appeared on any document: extended payment terms, free goods, absorbed freight, unbilled work, credits, rebates paid later, cooperative advertising allowances and anything else given to obtain the order. The pocket margin is the pocket price minus the unit cost of the offer, expressed in dollars or as a percentage of the pocket price. The distance between list price and pocket price is called the pocket price waterfall, a device introduced by Marn and Rosiello in 1992 to show how price erodes between a company's invoice figure and the amount the customer actually pays, and developed further in The price advantage (Marn & Rosiello, 1992; Baker et al., 2010). In plain terms: the list price is what you publish, the invoice price is what you admit, the pocket price is what you keep, and only the last one pays the wages.

6.1.3 The cost of a discount, in a form you can put on a wall

The following table, Table 6.1, sets out what each level of discount does to the kit in the running case, which is priced at $4,800 against a unit cost of $2,100, for a margin rate of 56.25 percent. Its columns give the discount off the list price, the price that results, the margin left in dollars, the share of the list margin given up, and the extra unit volume that would be needed to earn the same total margin as before. The last column is the one that changes behavior, because it converts a concession that sounds modest into a sales target that is plainly unreachable.

Table 6.1 — What each discount costs on a kit priced at $4,800 with a $2,100 cost

Two readings of that table matter. The first is the acceleration: the share of margin given up rises in a straight line, but the volume needed to replace it rises faster and faster, because each discount both removes margin and shrinks the margin that must do the replacing. The second is the floor: at the company's own margin floor of $3,500, nearly half the margin of the transaction is gone, and the business would need to sell almost twice as many units to be where it started. A floor is not a normal price. It is the last price before the answer becomes no.

6.1.4 The three kinds of discount, which behave very differently

The strategic pricing literature treats policy as a decision of its own, distinct from the structure of a price and from its level, and places it between the two: once the business has decided how the price is shaped, it must decide the rules under which that price varies from one buyer to another and from one moment to the next, and only then does the level matter (Nagle et al., 2023). This chapter is that middle decision, and it begins by observing that not every reduction is the same animal, and that a policy which treats them alike will fail. A structural discount is published, conditional and available to everyone who meets the condition: a volume step, an annual prepayment, a distributor price. It belongs to the price structure of Chapter 5, it is announced in advance, and it is earned by the buyer doing something specific. A promotional discount is temporary, applies to a defined offer for a defined period, and is aimed at a market rather than at a person; Section 6.6 designs it. A negotiated discount is discretionary, granted to one buyer in one transaction because that buyer asked; it is the only one of the three that has no condition attached unless somebody attaches one, and it is where most margin is lost.

The distinction does real work in a policy. A structural discount does not need approval, because the approval happened when the price list was published; it needs a fence, which is the condition that must genuinely be met. A promotional discount needs a calendar, a budget and a measurement. A negotiated discount needs an authority limit, a counterpart and a record. Confusing the three produces the two most common failures in small businesses: a structural discount granted to a buyer who does not meet the condition, which is simply a negotiated discount wearing a uniform, and a promotional discount that never ends, which is a price cut that nobody decided.

KEY CONCEPT A discount is a permanent transfer of margin bought with a temporary feeling The money a discount moves does not come back. It is transferred from the seller's margin to the buyer's, in full, on the day the order is signed, and it stays transferred for every renewal and every follow-on order, because the buyer's reference price is now the discounted one. What the seller receives in exchange, unless something specific was traded for it, is a feeling: relief that the negotiation ended, reassurance that the deal will close, the impression of having been competitive. The feeling lasts an afternoon and the transfer lasts for the life of the account. This is why the discipline of this chapter is not "discount less" but "know what you bought." A discount traded for a five-building commitment, a two-year term, a thirty percent deposit or a reference the business can publish is an investment with a return that can be computed. A discount granted to end a silence on the telephone is a gift, and a gift that the recipient will expect again.

6.2 The Pocket Price Waterfall

6.2.1 What the waterfall is, and why it exists

A pocket price waterfall is a simple accounting of everything that stands between the published price of one transaction and the money the business keeps from it. It was introduced as a management tool by Marn and Rosiello in their 1992 study of pricing as a profit discipline, precisely because the reductions that matter most are the ones no system records, and it remains the first exercise to perform when a business suspects it is discounting more than it knows (Marn & Rosiello, 1992). The same authors describe a companion device, the pocket price band, which plots the pocket prices actually obtained for the same product across all customers; the publisher's description of the article reports that a variation of more than sixty percent between the highest and lowest pocket price is commonly observed among manufacturers. A business that has never drawn its own band should expect to be surprised by its width.

The reason the exercise is revealing is structural rather than psychological. Accounting systems record what is on the invoice, because that is what they are for. A payment term, a free item, an unbilled hour and an absorbed delivery are not price reductions in the eyes of the ledger; they are, respectively, a financing decision, an inventory movement, a labor cost and an expense. Each lands in a different account, none of them is attributed to the transaction that caused it, and the gross margin rate of the business therefore falls by a little every year for reasons that no report explains.

6.2.2 The six lines of one order

The order described at the opening of this chapter can now be taken apart. The list price of one kit is $4,800. The buyer had five buildings, which meets the published volume step of five to nineteen buildings and earns 6 percent, or $288.00, bringing the invoice price to $4,512.00. That is a structural discount: it was published in advance, the buyer met the condition, and it is defensible. What followed was not published. A further 5 percent off the invoice price, $225.60, was conceded on the telephone at the end of a long negotiation, bringing the negotiated price to $4,286.40. The buyer then asked for sixty days to pay instead of the standard thirty; at the company's cost of money of 12 percent a year, thirty extra days on $4,286.40 costs $42.28. A pack of six spare sensors, which sells for $390 and costs the company $120, was added at no charge; the waterfall counts $120, because the pocket price measures money, and the money the company gave up is what the pack cost it, not what it would have sold for. Finally, the installation ran longer than estimated and the extra work, worth $150, was never billed.

The following table, Table 6.2, lays out those six lines in the order they occurred, with the amount of each, the price remaining after it, that price as a percentage of the list price, and — the column that explains why none of this was visible — where each reduction is recorded in a normal accounting system. Reading down the last column is the fastest way to understand why a business can be certain it discounts six percent while in fact it discounts more than seventeen.

Table 6.2 — The pocket price waterfall of one negotiated order, line by line

DO THE MATH The pocket price of one negotiated order Fictional example; figures are invented and computed by the case model. The list price of a kit is $4,800.00 and its unit cost is $2,100.00, so the list margin is $2,700.00, a margin rate of 56.25 percent. The published volume step for five to nineteen buildings takes 6 percent, or $288.00, leaving an invoice price of $4,512.00. A negotiated 5 percent off the invoice price takes a further $225.60, leaving $4,286.40. Thirty extra days of payment terms on $4,286.40, at a cost of money of 12 percent a year, cost $42.28, because $4,286.40 × 0.12 × 30 ÷ 365 = $42.28. A free pack of six spare sensors costs $120.00, which is what the company paid for it. An unbilled installation overrun costs $150.00. The pocket price is therefore $3,974.12, which is 82.79 percent of the list price, and the pocket margin is $1,874.12 against a list margin of $2,700.00 — 30.59 percent of the margin, gone. The three lines that never reached the invoice total $312.28, which is 6.51 percent of the list price on their own. The invoice, meanwhile, shows a discount of 6 percent.

The following figure, Figure 6.1, draws the same seven lines as a staircase falling from the list price to the pocket price, with each step labeled by its amount and each landing labeled by the percentage of the list price that survives it. Two dashed lines cross the chart: the margin floor of $3,500 set in Section 6.3, and the unit cost of $2,100, which is the distance between the last bar and the bottom of the chart that the business actually keeps.

Figure 6.1 — The pocket price waterfall of one negotiated order. A staircase of seven bars falls from a list price of $4,800.00 to a pocket price of $3,974.12, which is 82.79 percent of the list price. The five steps between them are a volume discount of $288.00, a negotiated discount of $225.60, the cost of extended payment terms of $42.28, a free sensor pack costing $120.00 and an unbilled installation overrun of $150.00. A dashed line marks the margin floor of $3,500 and a second dashed line marks the unit cost of $2,100. The pocket margin is $1,874.12 against a list margin of $2,700, so 30.59 percent of the list margin is gone, while the invoice records a discount of six percent only. Hydrosentinelle inc. is fictional.

6.2.3 What the waterfall proves

Three conclusions follow from that single order, and each one generalizes.

The first is that the visible discount is a poor estimate of the real one. The invoice records 6 percent; the total erosion is 17.21 percent of the list price. A business that manages discounting by watching invoice discounts is watching roughly a third of the problem.

The second is that the margin damage is always larger than the price damage, and by a predictable factor. The price fell by 17.21 percent and the margin fell by 30.59 percent, because the cost did not move at all. The ratio between the two is fixed by the margin rate: at a margin rate of 56.25 percent, every point of price is one and three-quarter points of margin.

The third is that the three invisible lines were never negotiated. Nobody sat down and decided to spend $312.28 on payment terms, spare sensors and unbilled labor; each of the three was a small accommodation made by a different person at a different moment, and together they cost more than the negotiated discount that the sales team argued about for two weeks. This is the ordinary shape of the problem in a small business: the discount that everyone debates is smaller than the sum of the concessions that nobody counts.

CASE IN POINT The six lines that took thirty percent of the margin Fictional example; names and figures are invented. When Laurence Bergeron, the president of Hydrosentinelle inc., rebuilt the March order line by line, the reaction in the room was not guilt but surprise: nobody present recognized the transaction. Karim Haddad, who leads sales and operations, had negotiated the 5 percent on the telephone and remembered it clearly, and had assumed it was the whole of the concession. The sixty-day payment term had been agreed by the same person in the same conversation and had never been thought of as money. The free sensor pack had been authorized by the technician who delivered the kits, on the reasonable ground that the buildings were larger than expected. The unbilled installation overrun had been decided by nobody at all: the crew stayed late, the hours went into the payroll system, and no one connected them to the order. Four people had each made a defensible small decision, and the pocket price was $3,974.12. The company's response was not to reprimand anyone. It was to write the policy of Sections 6.3 to 6.5, so that the next four defensible small decisions would be made by people who could see the first three.

6.2.4 The concessions that never look like money

The concessions that escape every system share one property: they are denominated in something other than dollars at the moment they are granted. Days, items, hours and promises all convert into money later, quietly, in another account. A discount policy that does not price them in advance will not control them, because the person granting them does not experience them as price at all.

WATCH OUT An extended payment term is a discount nobody wrote down Sixty days instead of thirty on a $4,286.40 invoice, at a cost of money of 12 percent a year, is $42.28 — small enough to feel like a courtesy and large enough, repeated across a year of orders, to matter. Four other concessions behave the same way and belong in the same list. Free goods cost what they cost you, not what they sell for, and the correct entry in the waterfall is the unit cost. Unbilled work — the overrun, the extra visit, the training session added at the end — is pure margin, because the labor was paid and nothing was collected for it. Absorbed freight or delivery is a price reduction expressed as an expense. And a rebate paid later, whether it is a volume bonus, a cooperative advertising allowance or a year-end credit, is a discount with a delay, which is the hardest kind to see because it is granted in one period and paid in another. Price every one of these in dollars before you grant it, write the dollars on the same line as the order, and the conversation with the buyer changes: a request for sixty-day terms becomes a request for forty-two dollars, and forty-two dollars can be traded for something.

6.3 The Margin Floor

6.3.1 A floor is a margin, not a discount percentage

The single most useful document a business can write about discounting is one number, and the number is a margin. Hydrosentinelle's floor is a minimum kit margin of 40 percent, which at a unit cost of $2,100 produces a floor price of $3,500, because $2,100 ÷ (1 − 0.40) = $3,500. That floor is 27.08 percent below the list price of $4,800 and leaves $1,400 of margin on every kit sold at it.

Expressing the floor as a margin rather than as a maximum discount matters for two reasons. The first is that costs move and prices do not, so a floor written as "never more than twenty-five percent off" silently becomes a different floor every time a supplier raises a price, and always in the wrong direction. A floor written as a margin rate recomputes itself. The second is that a margin floor forces the business to know its unit cost, which is the discipline that makes every other pricing decision possible; Book G1 (Accounting Fundamentals for Business Owners) computes that cost properly, and this chapter consumes the result.

The floor also has to be a real refusal. A floor that has been crossed twice is not a floor; it is a suggestion, and every person who negotiates knows which it is. The test of whether a business has a floor is not whether the number exists but whether anyone has ever lost a deal at it and been thanked for it.

6.3.2 The floor is measured on the pocket price, not on the invoice

Here is where the two halves of this chapter meet. A floor applied to the invoice price is not a floor at all, because the invoice is exactly the point at which the measurement stops being accurate. On the March order, the invoice price of $4,512.00 against a cost of $2,100 shows a margin rate of 53.46 percent, comfortably above the 40 percent floor. The pocket margin rate is 47.16 percent, because the pocket price is $3,974.12 and the margin is $1,874.12. The order still clears the floor, but by six and a third points less than the invoice suggested, and a second order with slightly more generous terms would clear it only on paper.

The practical consequence is that a floor has to be stated with the concessions included. Hydrosentinelle's policy therefore reads: no kit is sold below a 40 percent pocket margin, counting free goods at cost, extended terms at the company's cost of money, and any unbilled work at its labor cost. That sentence is longer than "no more than twenty-five percent off," and it is the only version that survives contact with a negotiation.

6.3.3 What sits below the floor, and who may go there

A floor needs one exception rule, written in advance, or it will acquire a hundred unwritten ones. The rule has three parts. First, only one person may authorize a sale below the floor, and in a small business that person is the owner or the president rather than anyone whose compensation depends on closing; in the case, that is Laurence Bergeron and no one else. Second, the authorization is per transaction and never per account, so that a below-floor price granted once does not become a standing arrangement. Third, the authorization is written with its reason and its expiry, and it is counted: a business that authorizes three below-floor transactions in a year has an exception, and a business that authorizes thirty has a price level problem that Chapter 2 should be settling instead.

There are genuine reasons to sell below a floor, and naming them prevents the list from expanding. Entering a segment with a first reference customer whose name may be published is one, provided the publication is written into the agreement. Absorbing a one-time cost that will not recur — a first installation in an unfamiliar building type — is another, provided the price for the second installation is agreed at the same time. Clearing inventory that is obsolete or perishable is a third, and it is the only one that has nothing to do with the buyer. Everything else is either a price level problem or a targeting problem, and Book E2 (Prospecting and Outbound Sales), Chapter 2, treats the second.

6.3.4 The floor in a service business, and the floor for a subscription

Two adaptations are worth stating, because most readers of this book do not sell a kit with a clean unit cost.

In a business that sells time, the floor is a minimum effective hourly rate rather than a margin rate, computed as the fee divided by the hours the work genuinely takes, including the unbilled ones. The discipline is the same and the arithmetic is easier to abuse, because the hours are estimated by the person who wants the deal. The remedy is to compute the effective rate after delivery, on the hours actually recorded, and to compare it with the floor at that point; a business that does this for one quarter discovers which of its offers are quietly below the line.

In a subscription business, the floor has a second dimension, which is time. A monthly price discounted at the start is discounted for the life of the account unless the increase is written into the agreement, so the floor must be expressed as the margin over the expected life rather than over the first month. Hydrosentinelle's monitoring subscription costs $15 a month to serve and sells for $79; a promotional first year at $59 would still be profitable in the month, at a margin rate of 74.58 percent, and would nonetheless be a bad trade if the account never returns to $79, because the reference price of that customer is now fifty-nine dollars and every renewal conversation starts there. Chapter 8 treats the increase that such an arrangement requires, and Book H1 (Contracts and Commercial Agreements) drafts the clause that makes it enforceable.

6.4 Discount Authority and Approval

6.4.1 Authority is the mechanism that makes a floor operate

A floor tells a business where to stop. Authority tells each person how far they may travel on their own, and it is what converts a policy into a daily practice. Without it, every negotiation is conducted at the floor by default, because the person in the room has no intermediate stopping point and no reason to invent one.

Hydrosentinelle's ladder has four rungs and is deliberately short. A sales representative may grant up to 6 percent off the list price without asking anyone. The head of sales may grant up to 12 percent. Beyond twelve percent and down to the floor, the president decides. Below the floor of $3,500, nobody decides without the exception procedure of Section 6.3.3. The two intermediate limits are not arbitrary: they are set at the company's own published volume steps, so that a discretionary discount can never quietly exceed the largest discount the price list itself offers. A representative who grants six percent to a single-building buyer has handed out the five-building price, and everyone in the company can see that on one page.

The following table, Table 6.3, states the ladder as it would be written in the policy. Its columns give the role, the maximum discount off the list price, the price that discount reaches on a kit, and what the person must record when they use it. The recording column is the one that is usually missing, and its absence is what makes the rest of the table unenforceable.

Table 6.3 — Who may approve what, on a kit listed at $4,800

6.4.2 Approval has to be faster than the deal

The most common reason a discount policy fails is not that people disagree with it; it is that asking takes longer than conceding. If a representative must wait two days for an answer, the policy will be circumvented by people acting in good faith to save a sale, and it will be circumvented in the invisible currencies of Section 6.2.4, where no approval is required at all. An approval step must therefore have a stated turnaround — same day is a reasonable standard in a small business — and a default. A default of "no" is safer than a default of "yes," and either is better than silence, because silence teaches people to stop asking.

The second most common reason is that the approval asks the wrong question. "May I give eight percent?" invites a yes or a no about a number. "The buyer wants eight percent; I have offered the twenty-four-month term and the thirty percent deposit in exchange, and they have refused the term" invites a decision about a trade. The request form should make the second question the only one that can be asked, by requiring the counterpart field to be filled before the request can be sent.

WHO DOES THIS The representative, the head of sales and the owner, and why the last word is never the seller's In a solo business, one person plays all three roles, and the separation must be made in time rather than in people: the floor and the authority limits are written on a scheduled day, in writing, and treated as binding during a negotiation exactly as an employee would treat them. The owner who wrote $3,500 on Monday is the authority that the owner on Thursday must ask, and the answer on Thursday is given by the document. In a small team, the person who owns the margin and the person who closes the deal should not be the same person, and in Hydrosentinelle inc. the price belongs to Laurence Bergeron, the president, while Karim Haddad, who leads sales and operations, holds an authority limit of 12 percent and a written list of acceptable trades. In an established business with several sellers, the ladder becomes a control that must be audited rather than trusted: someone reads the discounts granted each quarter, by person and by account, and the reading is part of the sales management cadence that Book E8 (Sales Management: Pipeline, Forecasting, Coaching and Compensation) describes. In every configuration the rule is the same, and it is the reason the ladder exists: the last word on a price never belongs to the person whose income depends on the deal closing.

6.4.3 What the policy has to contain, and how long it should be

A discount policy that runs beyond two pages will not be read, and a policy that is not read has no effect on behavior. The content that matters is short, and most of it has already been assembled in this chapter.

METHOD Writing a one-page discount policy When to use it: as soon as more than one person can adjust a price, and again at every annual price review, before the price list is republished. Steps: (1) Write the current list price of each main offer and the date it takes effect, so that every later reduction has a reference. (2) Compute the unit cost of each offer from the accounting records rather than from memory, and write the minimum margin rate the business will accept; convert it into a floor price, as the case does with a 40 percent minimum margin, a $2,100 cost and a floor of $3,500. (3) State that the floor is measured on the pocket price, and list the concessions that must be priced into it: free goods at cost, extended terms at the cost of money, unbilled work at labor cost, absorbed freight, and rebates paid later. (4) Write the authority ladder as a four-row table giving each role, its maximum discount, the price that reaches, and what must be recorded. (5) Write the list of acceptable counterparts, with the discount each one is worth, so that a seller trades from a menu rather than from imagination. (6) State the approval turnaround and the default answer if no decision arrives in time. (7) State what is recorded for every reduction: the amount, the counterpart, the author, the date and the approver. (8) Name the person who reads the record each quarter, and the date of the next review. Deliverable: one page, dated, signed by the owner, held where the people who negotiate will actually look, and given to every new seller on the first day. Verification criterion: a seller who has read it once can state, without looking, the floor price of the main offer, their own authority limit and three counterparts they are allowed to trade for a discount.

TOOLS OF THE TRADE What a discount approval workflow must enforce A workflow is judged on what it refuses, not on what it displays. A list price with an effective date, so that a discount is computed against a published reference rather than against last quarter's habit. A computed floor, derived from the recorded unit cost rather than typed in by hand, so that it moves when the cost moves. A hard stop, meaning that a price below the floor cannot be issued at all without a named exception, rather than merely displaying a warning that everyone learns to click through. A mandatory counterpart field, which prevents a request from being submitted until the seller has stated what is being obtained in exchange. Off-invoice capture, so that payment terms, free goods, unbilled work and freight are entered on the order and priced automatically at the company's cost of money and unit costs. A role-based limit that reflects the authority ladder and refuses beyond it instead of notifying somebody afterwards. A timer with a default, so that an unanswered request resolves the same day in a stated direction. An immutable log of who asked, who approved, what was traded and when, readable later by someone who was not in the room. A quarterly extract of every reduction granted, by seller, by account and by size, which is the raw material of the discount-band analysis of Chapter 9. Chapter 12 maps these functions against a spreadsheet, a market tool and La Machine, and says plainly which of them each one covers.

6.5 The Rule of Exchange: Never Give, Always Trade

6.5.1 The rule, and why it is a rule rather than a preference

Every discount is exchanged for something, or it is not granted. That sentence is the whole of the policy, and it is worth stating in the negative as well: the business does not reduce a price because a buyer asked, because a competitor is cheaper, because the quarter is ending, or because the silence on the telephone became uncomfortable. It reduces a price when the buyer gives it something the business wants, and the value of that something can be stated in dollars before the reduction is agreed.

The rule has three effects, and only the first is financial. It preserves margin, obviously. It also preserves the price level, because a discount attached to a condition is not a statement that the price was too high; it is a statement that a different transaction has a different price, which is true and defensible. And it changes the negotiation itself: a buyer who must give something to get something stops asking for reductions that are worth less to them than the counterpart, which means the requests that survive are the ones worth discussing. Book E1, Chapters 7 and 8, conducts that conversation; this chapter supplies the menu it negotiates from.

6.5.2 The counterparts worth having, and what each is worth

A trade is only real if the business can say what it received. The following table, Table 6.4, lists the counterparts a small business can realistically obtain, what each one is worth to the seller, and where it has to be written down so that it survives the enthusiasm of the moment. Its columns give the counterpart, why it has value, an indicative amount of discount it justifies in the case, and the document that records it. The third column is indicative rather than prescriptive: each business computes its own from its own margin and cost of money, by the arithmetic of Sections 6.1 and 6.2.

Table 6.4 — What a discount may be traded for, and where the trade is written

Three of those rows deserve a word of caution. A reference is worth a great deal and is very often not delivered, so the permission to publish should be signed at the same moment as the discount, never promised for later. Reduced scope is the only counterpart that genuinely lowers the seller's cost, which makes it the safest trade in the table and the most underused. And a faster decision is worth less than sellers believe: a buyer who signs two weeks earlier has given up almost nothing, so the step traded for it should be the smallest on the menu.

6.5.3 The annual prepayment, priced as what it is

The clearest example of a trade that must be computed rather than assumed is the discount for annual payment, which Book B1 (Marketing Fundamentals for Business Owners), Chapter 6, names as a signal that trades margin for cash and commitment, and leaves for this book to quantify. Hydrosentinelle offers 10 percent off for a year paid in advance: twelve monthly payments of $79 are $948.00, and the annual prepayment is $853.20, so the discount is $94.80.

What does that buy? It buys the use of money earlier. If the customer pays at the start of each month, the average monthly payment arrives five and a half months before the end of the year, so the prepayment brings forward the equivalent of $948.00 × 5.5 ÷ 12 = $434.50 of money for one year. At the company's cost of money of 12 percent, borrowing that same amount for a year would cost $52.14. The business is therefore paying $94.80 for cash that its own lender would supply for $52.14, which is an implicit annual rate of 21.82 percent, because $94.80 ÷ $434.50 = 0.2182. Judged as financing alone, the annual prepayment discount is expensive money. Section 4.3.5 reaches the same conclusion by the other route, comparing what the twelve monthly payments are worth today against what the prepayment brings in; the difference between the two figures is the convention used for when the money arrives, which each section states, and not a disagreement about the trade.

Judged as a retention instrument, it may still be worth it, and this is where a reader must weigh evidence carefully. A billing study published by ChartMogul, covering more than 2,500 subscription software companies for the full year 2024, reports that among companies with average revenue per account between $250 and $500, annual plans showed net revenue retention of 88 percent against 76 percent for monthly plans, and that at low revenue per account customer retention was 62 percent on annual plans against 41 percent on monthly ones (Faustino, 2025). That is a substantial difference, and it is the reason the trade is often correct. It is also, and the point must be made, a non-probabilistic sample of one vendor's own customers, aggregated from billing data, with no peer review, and drawn from software subscriptions rather than from building services. It supports a hypothesis about your business; it does not measure your business. The honest conclusion for the case is that the company pays $94.80 for financing worth $52.14, that the excess of $42.66 is what it is really spending on commitment and retention, and that the excess is small enough to be worth spending only if the retention effect is real — which the company must measure on its own accounts rather than borrow from a report. Chapter 4 treats the subscription model itself, Chapter 9 sets out how to measure the retention effect, and Book G2 (Cash Flow Management and Forecasting) manages the cash that arrives.

6.5.4 Curing systematic discounting

The symptom that brings most readers to this chapter is not one bad deal; it is that every deal closes below list. Book E1, Chapter 14, sends readers here precisely on that symptom. It has to be treated as a system failure rather than as a discipline failure, because blaming the sellers produces three months of compliance and then a return to the previous state.

The cure has four moves, in order. The first is measurement, not exhortation: compute the realized price for the last twelve months, as total pocket value divided by total list value, and read the distribution of discounts by band. On the case, Chapter 9 performs exactly that reading and finds a realization of 94.21 percent on the invoice and 91.21 percent in the pocket, with the most profitable band, measured per quote sent, being the one to five percent band rather than the deepest one. That single finding is more persuasive than any policy, because it shows the sellers that the deep discounts did not buy the win rate they were supposed to buy. The second move is the floor, because until a floor exists the answer to "how low can we go" is "lower." The third is the authority ladder, which removes the need for each seller to invent a stopping point. The fourth is the counterpart menu of Table 6.4, which gives the seller something to say other than yes or no.

There is a fifth move that applies when the first four have been made and the discounts persist, and it is uncomfortable: the list price may be wrong. A business in which every buyer negotiates the same reduction has published a price that the market does not accept, and it should either lower the list price and stop discounting, or fix the value argument that Chapter 2 supplies. A permanent, universal discount is not a discount. It is the real price, printed in the wrong place.

6.6 Promotions: Design, Calendar and Measurement

6.6.1 What a promotion is for, and the three legitimate reasons to run one

A promotion is a temporary reduction, applied to a defined offer for a defined period, addressed to a market rather than to a person. It differs from a negotiated discount in that nobody asked for it, and from a structural discount in that it ends. That ending is the whole of its logic: a promotion that does not end is a price cut that no one decided, and it destroys the reference price that made the promotion attractive in the first place.

There are three sound reasons to run one and several common bad ones. The sound reasons are to acquire trial in a segment that does not yet know the offer, where the promotion is an acquisition cost paid once against a customer lifetime that Chapter 9 measures; to clear something perishable, seasonal or obsolete, where the alternative is a write-off; and to fill capacity that would otherwise be idle in a known trough, where the marginal cost of serving is genuinely low. The bad reasons are to hit a quarterly number, which borrows revenue from the next quarter at a discount; to respond to a competitor's promotion, which begins a sequence that ends with both businesses poorer; and to reward loyal customers, which sounds generous and in practice hands a reduction to the people who were going to buy anyway.

6.6.2 A promotion is a budget, and the budget is measured in extra units

The arithmetic of a promotion is the arithmetic of Section 6.1 applied to a whole campaign. If Hydrosentinelle ran a fifteen percent promotion on the kit, the price would be $4,080.00, the margin $1,980.00, and the company would need 36.36 percent more units during the campaign merely to earn the same margin as it would have earned without it. That number has to be written down before the campaign starts, because after it starts everyone will look at revenue instead.

The measurement that matters is not units sold but incremental units sold: the units that would not have been sold without the promotion. Every unit that would have been bought anyway is a unit sold at fifteen percent off for no reason, and in most promotions those units are the majority. The practical estimate available to a small business is a comparison with a matched period — the same weeks of the previous year, or the immediately preceding weeks, adjusted for anything obviously different — and the estimate should be written down with its assumption stated. It will be imperfect. It will still be far better than counting revenue, which always rises during a promotion and proves nothing.

The following table, Table 6.5, gives the elements a promotion must state before it runs and the question each element settles, with the case's fifteen percent kit promotion as the worked example. Its three columns run from the element to the example, and it is meant to be used as a form: a promotion whose form has blank rows is not ready to run.

Table 6.5 — What a promotion states before it runs

6.6.3 The calendar, and the trap of the permanent sale

A promotional calendar exists to answer one question in advance: how often may we do this? A business that promotes continuously teaches its market to wait, and the waiting is rational: a buyer who knows that a reduction arrives every six weeks would be foolish to pay the list price. The damage is measurable on the margin and, as Section 6.7 explains, it is also a legal problem, because a price that is rarely charged stops being the ordinary price and a comparison to it stops being defensible.

The American guide on the word "free" offers a discipline that is worth borrowing even where it does not apply. It states that a "free" offer on a given item should not run for more than six months in any twelve, that at least thirty days should separate two such offers, that no more than three should run in the same trade area in twelve months, and that sales made under the offer should not represent more than half of the product's total volume. Those thresholds are a United States guide on one kind of offer, not a Canadian rule and not a general rule about discounts; but a small business that adopts them as its own calendar discipline will never accidentally turn its list price into a fiction.

6.6.4 The mechanics that make a promotion controllable

Four mechanics separate a promotion that can be stopped from one that cannot. A published end date, which is also the date the system reverts, rather than a date someone must remember to act on. A cap, whether on units, on total discount dollars or on the number of customers, so that a promotion that works far better than expected does not become the year's largest unplanned expense. A qualification condition that can be checked, such as a new building or a first order, so that existing customers do not receive a reduction they were not going to ask for. And a record of the ordinary price in force before and after, because the evidence that a comparison was truthful must exist before anyone asks for it, and cannot be reconstructed afterwards.

IN LA MACHINE A promotion is a whole-number percentage, a period and a list of items In La Machine, a promotion (business/promotions, also reachable at offers/promotions, and visible to the customer at client/promotions) holds a name, a tagline, a description, an image, translations, a whole-number percentage from 0 to 100, a status of draft, active or expired that a person sets by hand, a start date, an end date, and a list of items it applies to, which may be a subscription, a session package or a store product. Sending a promotion creates internal alerts and writes send records; the email itself is a simulated log entry rather than a message. On the customer's side, the reduced price is displayed. That is the whole of the feature, and it is the two things this section asks for most: a percentage and a period. What is not here: there is no promotion code and no coupon, no fixed-dollar amount, no usage limit, no minimum order, no stacking rule, no free month and no credit, and — the point that matters most — no promotion is applied at payment, because the checkout is a demonstration in which nothing is charged. There is also no discount approval, no margin floor, no authority by role and no record of a discount granted outside a promotion. A business that runs its promotions here still computes the break-even, keeps the ordinary-price evidence and controls the calendar somewhere else.

6.7 The Law of Promotional Prices

6.7.1 The principle that both countries share

Before any statute is cited, the principle common to Canadian and American law can be stated in three sentences, and a business that follows them will rarely be in difficulty on either side of the border. The first price a buyer sees must be a price the buyer can actually obtain, with only charges imposed by law added afterwards. A price presented as a former price, or as the usual price, must be a price at which the seller genuinely offered or sold the thing, over a period long enough to mean something, and the seller must be able to prove it. And a seller must supply at the price it advertised.

Book B1, Chapter 6, states the practical version of that principle for the whole library, and this book keeps it word for word: the price the buyer sees first is the price they pay, apart from taxes, and every "was" price is real and documented. What follows adds the provisions, because a rule a business can cite is a rule a business can defend.

6.7.2 The Canadian provisions, one by one

Five provisions of the Competition Act govern a promotional price, and they operate in two regimes. Part VII.1 treats conduct as reviewable, which is civil: no intention has to be proved, and the remedies are orders, restitution and administrative monetary penalties. Section 52 treats the same representation as a criminal offense when it is made knowingly or recklessly, which is why the difference between the two regimes is a state of mind and not a category of conduct.

Section 74.01(1.1) makes it reviewable conduct to display a price that is not attainable because of fixed obligatory charges, with an exception only for amounts imposed by an Act of Parliament or of a provincial legislature, which in practice means sales taxes. Subsections 74.01(2) and 74.01(3) govern the ordinary price claim, which is the provision that most promotions engage: subsection (2) applies when the comparison is to the price at which suppliers in general supply the product in the relevant geographic market, and subsection (3) applies when the representation is clearly about the advertiser's own usual price. In both cases the claim is reviewable unless one of two conditions is met — a substantial volume of the product was sold at that price or more within a reasonable period, or the product was offered in good faith at that price or more for a substantial period of time. The two conditions are disjunctive: satisfying either one is enough.

The figures everyone quotes for those two conditions are not in the statute, and attributing them to it is a mistake that survives audit. The Competition Bureau's publication on ordinary price claims sets out what it regards as satisfying them: more than fifty percent of sales at that price or higher in the twelve months before or after the representation for the volume test, and the product offered in good faith at that price or higher for more than half of the six months before or after for the time test, with an adjusted period for seasonal products (Competition Bureau Canada, 2022). That is enforcement policy published by the administering agency, which is authoritative about how the Bureau will proceed and is not the text of the law.

Section 74.05 makes it reviewable conduct to advertise a product for sale or rent in a market and then supply it, during the period and in the market covered by the advertisement, at a price above the advertised one. There are narrow exceptions: a catalog carrying a prominent statement that its prices may contain an error, where the advertiser proves the error; an advertisement immediately followed by a correction; securities; and a product supplied by someone not in that business. Section 52 makes the knowing or reckless version of a false or misleading representation a criminal offense, and subsection 52(4) provides that the general impression conveyed is taken into account as well as the literal meaning, so a technically accurate promotion whose overall impression misleads is still caught. Section 74.1 sets the penalties for the civil side, and they are not small for a small business: for a corporation, a first order may carry the greater of ten million dollars and three times the benefit derived from the conduct, or, where that benefit cannot reasonably be determined, three percent of annual worldwide gross revenues; for an individual, the greater of $750,000 and three times the benefit. Subsection 74.1(3) provides a due diligence defense, which is the legal reason a documented promotion process is worth maintaining.

QUÉBEC & CANADA NOTE An advertised reduction must be real, and the register must match the shelf A business selling to consumers in Québec is subject to a second regime, stricter about display than the federal one. The Consumer Protection Act, CQLR c P-40.1, section 224, provides that a merchant may not, in an advertisement, give less prominence to the total price of an assembly of goods or services than to the price of a part of it, may not disclose the amount of installments without disclosing and giving prominence to the total price, and may not charge a higher price than the advertised price; that is the reading made at source for Book E1 on September 8, 2026, at the consolidation of April 7, 2026, and LégisQuébec returned an access error on September 19, 2026, so the text was not read again for this book. The Office de la protection du consommateur, which administers the Act, states on its page of February 10, 2025 that the advertised price must equal the total the customer has to pay, that administration, transport and preparation fees must be included, that the GST, the QST and specific contributions remitted to a public authority may be added at payment, and that announcing a reduction that does not exist is prohibited (Office de la protection du consommateur, 2025). A separate rule applies in stores. Under the Price Accuracy Policy, which binds merchants exempted from labeling each item individually, a good scanned at the register for more than its displayed price is free when the displayed price is $15 or less, and otherwise is sold at the displayed price less $15; the policy applies in store and not online, and the compensation is computed from the lowest displayed price (Office de la protection du consommateur, 2026). Hydrosentinelle inc. sells to building owners and managers, so most of its transactions are commercial rather than consumer transactions and the Consumer Protection Act does not govern them; the Competition Act governs them all the same. Verify which regime applies to your own buyers before relying on either.

6.7.3 What a reader who sells into the United States must know

The American rules on promotional pricing are built differently, and the difference is worth one paragraph before the comparison. The Guides Against Deceptive Pricing, 16 C.F.R. part 233, are interpretive guides issued in 1967 under section 5 of the FTC Act: they state what the Commission regards as deceptive, and they do not create a civil penalty of their own. A trade regulation rule, by contrast, carries civil penalties; the Rule on Unfair or Deceptive Fees, 16 C.F.R. part 464, effective May 12, 2025, is such a rule, and it requires the total price to be disclosed more prominently than any other pricing information — but it covers live-event tickets and short-term lodging only, so most readers of this book are governed by the guides and by section 5 itself rather than by that rule. Treating a guide and a rule as equivalent is the most common error in cross-border compliance summaries.

The following table, Table 6.6, compares the two countries on the six practices a discount and promotion policy has to settle. Its columns give the practice, the Canadian position, the American position, and the practice that satisfies both, which is the column to adopt if the business sells on both sides of the border. Every provision named in the table is listed in full in the chapter references, and the table states positions rather than advice: a transaction that matters should be checked against the current text and, where the amounts are significant, with counsel.

Table 6.6 — Discount and promotion practices, Canada and the United States compared

WHAT CHANGED Since 2022, the fine for price-fixing has no ceiling, and the civil penalties for a misleading price are far larger Before the 2022 amendments to the Competition Act, the maximum fine for the criminal conspiracy offense of section 45 was a fixed dollar ceiling, and it is still quoted as twenty-five million dollars in a great many secondary sources. It is no longer the law. In the consolidation current to July 21, 2026 and last amended March 26, 2026, read on September 19, 2026, subsection 45(2) provides that a person convicted on indictment is liable to imprisonment for up to fourteen years and to a fine in the discretion of the court, with no stated maximum. On the civil side, section 74.1 now sets administrative monetary penalties for reviewable conduct at, for a corporation, the greater of ten million dollars and three times the benefit derived, with a fallback of three percent of annual worldwide gross revenues where the benefit cannot reasonably be determined, and higher figures for a subsequent order. For a business planning a promotion, the practical consequence as of September 19, 2026 is that the old reassurance that a misleading price claim is a minor administrative matter is obsolete, and that the due diligence defense of subsection 74.1(3) — a documented process, kept evidence and a person responsible — is worth more than it used to be.

6.8 What You May Never Agree with a Competitor or Impose on a Reseller

6.8.1 The one conversation that is never worth having

Section 45 of the Competition Act makes it a criminal offense to conspire, agree or arrange with a competitor to fix, maintain, increase or control the price for the supply of a product, to allocate sales, territories, customers or markets, or to fix, control, prevent, lessen or eliminate the production or supply of a product. The offense is complete on the agreement: no effect on competition has to be proved, and the conduct is prohibited in itself. Subsection 45(3) provides that a conspiracy may be inferred from circumstantial evidence, and subsection 45(8) contains the definition that makes this a matter for a chapter on discounting rather than a matter for a chapter on strategy: for the purposes of the section, price includes a discount, rebate, allowance, price concession or other advantage in relation to the supply of a product. An agreement among competitors to stop offering a rebate, to standardize a payment term, or to hold a minimum discount is an agreement about price, and it is criminal.

The American position is the same in substance and arrives by a different route. Section 1 of the Sherman Act declares every contract, combination or conspiracy in restraint of trade illegal, and the penalties in the text are a fine of up to one hundred million dollars for a corporation, one million dollars for an individual, or up to ten years of imprisonment. The per se character of horizontal price fixing — the principle that no justification, however reasonable the resulting prices, is a defense — is not in the text of section 1; it comes from the case law, and the Antitrust Division of the Department of Justice states it in its primer for businesses, which also lists the forms such an agreement takes: agreeing on discounts, eliminating or reducing discounts, adopting a standard formula for computing prices, maintaining price differentials between quantities or grades, fixing credit terms, and agreeing not to advertise prices (U.S. Department of Justice, Antitrust Division, n.d.). The same primer notes that no written agreement is required and that proof is usually circumstantial.

The practical rule for a small business is therefore short and absolute. Prices, discounts, terms, margins, bids and customers are never discussed with a competitor: not at a trade association meeting, not at a conference dinner, not in a message that seems harmless, not "just to understand the market." If a competitor raises the subject, the correct response is to say plainly that the conversation cannot continue, to leave, and to write a note of what happened and when. Chapter 11 returns to the boundaries of an industry association, which are narrower than most owners assume.

6.8.2 What a supplier may and may not do to a reseller's price

The second prohibition catches businesses that sell through others, and it is far less known than the first. Section 76 of the Competition Act governs price maintenance. It was a criminal offense until 2009 and is now a reviewable practice before the Competition Tribunal, open to the Commissioner and, with leave, to a private party. Unlike section 45, it has two limbs that must both be established: the conduct, and an adverse effect on competition in a market.

THE RULE Competition Act, s. 76: a supplier may not push a reseller's price up or punish it for pricing low Jurisdiction: Canada, federal. Instrument: Competition Act, R.S.C. 1985, c. C-34, s. 76, in the consolidation current to July 21, 2026 and last amended March 26, 2026, read on the Justice Laws Website on September 19, 2026. Subsection 76(1) allows the Tribunal to make an order where a person, directly or indirectly, has by agreement, threat, promise or any like means influenced upward, or discouraged the reduction of, the price at which its customer or another person supplies, offers to supply or advertises a product in Canada, or has refused to supply a product to, or otherwise discriminated against, a person because of that person's low pricing policy — and where the conduct has had, is having or is likely to have an adverse effect on competition in a market. Both limbs are required. Subsection 76(5) provides that suggesting a resale price is itself proof of influencing upward unless the supplier makes it clear to the person that they are under no obligation to accept the suggestion and would in no way suffer in their business relations if they did not accept it. Subsection 76(6) applies the same logic to a minimum price advertised by someone who is not the retailer, unless the advertisement makes clear that the product may be sold at a lower price. Subsection 76(2) allows the Tribunal to prohibit the conduct or to order the supplier to accept the person as a customer on usual trade terms, which subsection 76(12) defines as terms of payment, units of purchase and reasonable technical and servicing requirements. The Bureau's enforcement guidelines add that the adverse-effect threshold is lower than the substantial-lessening test of other provisions, that without market power there is generally no adverse effect, and that below a market share of thirty-five percent the Bureau will not generally examine market power further — an enforcement policy, and not an immunity (Competition Bureau Canada, 2014). Scope for a reader of this book: any business that sells through dealers, distributors, agents, franchisees or online resellers, and any business that publishes a suggested retail price. Verify the current text before relying on it.

What that leaves open is more useful than what it forbids. A supplier may publish a suggested retail price, provided the freedom of the reseller is stated clearly, in writing, on the document that carries the suggestion. A supplier may set its own wholesale prices and its own volume steps. A supplier may decline to supply for reasons unconnected to the reseller's pricing, and may set genuine requirements of service, training or presentation. What it may not do is make the reseller's price a condition of supply, or punish a reseller that discounts, and the punishment does not have to be explicit: slower deliveries and withdrawn support are the forms this usually takes, and they are the forms most easily proved from a paper trail.

6.8.3 Different prices for different buyers, which is allowed and must be structured

Owners frequently assume that charging two buyers different prices for the same thing is unlawful. In the provisions read for this book, it is not: Canada's price maintenance provision reaches a supplier that discriminates against a reseller because of its low pricing policy, and none of the provisions examined here prohibits differential pricing as such. In the United States the position is different for goods, because the Robinson-Patman Act makes it unlawful to discriminate in price between purchasers of commodities of like grade and quality where the effect may be substantially to lessen competition, with a cost-justification defense for differences that reflect genuine differences in the cost of manufacture, sale or delivery, and a defense for a lower price offered in good faith to meet — not to beat — an equally low price of a competitor. The Act also makes it unlawful for a buyer knowingly to induce or receive such a discrimination, which is a point that large purchasers' procurement departments do not always volunteer.

The practical conclusion is the same on both sides of the border, and it is the conclusion this chapter has been building toward from Section 6.1. Differences in price should be published, conditional and available to anyone who meets the condition, and the condition should correspond to something real: a volume, a term, a delivery method, a scope, a channel. A published volume step of six percent for five to nineteen buildings is a defensible price structure with a documented cost rationale. The same six percent, granted informally to one buyer who asked and refused to another who did not, is the beginning of a problem that is at once commercial, evidentiary and legal. Chapter 5 builds the structure; this chapter keeps the discretionary part of the price small enough to govern.

Costly Mistakes

• Quoting a discount as a percentage of the price instead of a percentage of the margin. Ten percent off a kit sounds like a tenth of something; it is 17.78 percent of the margin and would take 21.62 percent more volume to replace.
• Managing discounts by watching the invoice. On the case, the invoice showed six percent while the real erosion was 17.21 percent of the list price and 30.59 percent of the margin, because three of the six reductions were denominated in days, items and hours rather than in dollars.
• Writing the floor as a maximum discount percentage. A floor of "no more than twenty-five percent off" moves every time a supplier changes a price; a floor of "no less than a forty percent pocket margin" recomputes itself and forces the business to know its unit cost.
• Granting a discount without naming the counterpart. A reduction traded for a volume commitment, a term, a deposit or a publishable reference is an investment; the same reduction granted to end a silence is a permanent transfer of margin in exchange for an afternoon of relief.
• Running a promotion with no end date, no cap and no break-even. A promotion that never ends becomes the real price, destroys the reference price that made it work, and makes any later comparison to the old price indefensible under the ordinary price provisions.
• Discussing prices, discounts or terms with a competitor. Section 45 defines price to include discounts, rebates, allowances and price concessions, the offense is complete on the agreement, and the fine has been at the discretion of the court since the 2022 amendments.

Key Takeaways

• A discount comes entirely out of margin, so its real size is the discount divided by the margin rate, and the volume required to replace it rises faster than the discount itself.
• The pocket price is the only price that pays wages, and the distance between list price and pocket price is where a small business loses margin it never decided to spend.
• On the case, six lines took the price from $4,800 to $3,974.12, which is 82.79 percent of list and 30.59 percent of the margin, while the accounting system recorded a discount of six percent.
• A floor is a margin, it is measured on the pocket price, and it is only a floor if a deal has been lost at it; the case's floor of $3,500 comes from a 40 percent minimum margin on a $2,100 cost.
• Authority by role is what makes a floor operate day to day, and in the case it is 6 percent for a representative, 12 percent for the head of sales, the president down to the floor, and a written exception below it.
• Every discount is traded for something priced in advance, and the counterpart that is safest is the one that genuinely lowers the seller's cost, such as reduced scope or self-installation.
• The annual prepayment discount is expensive financing and may still be a good trade: $94.80 buys money the company could borrow for $52.14, an implicit rate of 21.82 percent, and the retention it buys must be measured rather than assumed.
• A promotional price is a representation, so the former price must be genuine and documented, the advertised price must be honored, and the price the buyer sees first is the price they pay, apart from taxes.

Your Action Plan

1. Build one pocket price waterfall. The deliverable is a single sheet for one recent transaction, listing the list price, every reduction in the order it occurred with its amount in dollars, the pocket price and the pocket margin; it takes about ninety minutes with the invoice, the delivery notes and the time records in front of you, and it is done when the last line reconciles to the cash actually received minus the cost actually incurred.
2. Write your discount ladder in margin terms. The deliverable is a five-row table like Table 6.1 showing, for your main offer, what each level of discount leaves as margin and how much extra volume would be needed to replace it; it takes about thirty minutes once you have a reliable unit cost, and it is done when everyone who negotiates has a copy.
3. Set the floor. The deliverable is one number: the minimum margin rate you will accept, and the floor price it produces at your current cost; it takes about an hour, and it is done when the floor is written, dated, signed by the owner and stated to be measured on the pocket price with free goods, terms and unbilled work priced into it.
4. Write the authority ladder. The deliverable is a four-row table naming each role, its maximum discount, the price that reaches and what must be recorded, plus the approval turnaround and the default answer; it takes about forty-five minutes, and it is done when each person who negotiates can state their own limit from memory.
5. Build the counterpart menu. The deliverable is a list of six or seven counterparts, each with the discount it justifies and the document that records it, modeled on Table 6.4 but priced from your own margin and cost of money; it takes about an hour, and it is done when a seller can point to a row instead of inventing a concession.
6. Audit your promotional evidence. The deliverable is a folder containing, for every comparison price you currently display, the dated price list and the sales records that support it; it takes about two hours, and it is done when every crossed-out price in your marketing is supported by evidence you could hand to a regulator without editing it.
7. Schedule the quarterly discount read. The deliverable is a recurring quarterly task, with a named owner, that extracts every reduction granted in the quarter by person, by account and by size, and compares realized price to list; it takes about twenty minutes to set up and half a day each quarter to perform, and it is done when the first extract exists and has been discussed.

In La Machine

Every route named below was verified in the route inventory on September 19, 2026 and read in the product's source code at commit be8bf04 of September 7, 2026 (La Machine, 2026); nothing here is inferred from a page title.

• business/promotions (also reachable at offers/promotions) holds a promotion with a name, a tagline, a description, an image, translations, a whole-number discount percentage from 0 to 100, a status of draft, active or expired that a person sets by hand, a start date, an end date, and a list of target items that may be a subscription package, a session package or a store product.
• Sending a promotion writes internal alerts and send records; the customer email is a simulated log entry rather than a message that leaves the system. On the customer's side, client/promotions displays the reduced price.
• Business → Session Packages (business/packages, also productivity/packages) and business/subscription-packages hold a price and a cost on the same record, and the session package page computes profit per sale and margin percentage on screen without storing either; those two cost fields are the only place in the platform where the margin a discount would consume can be seen at all.
• Business → Store → Products (business/shop and business/shop-product-edit) holds a product cost, a shipping cost and shipping fees beside the price, and displays a profit panel computed from them; the comparison price — the crossed-out price — is displayed but is not saved, so the platform holds no evidence of a former price.
• Account → Profile → Revenues & Expenses (business/subscriptions) is the only view that places a sale's revenue beside its cost lines, and Control center → Audit log records who changed what, which is the nearest thing available to a trace of a price decision.
• What is not here. There is no discount approval, no margin floor, no authority by role and no price-change request of any kind. There is no promotion code and no coupon, no fixed-dollar reduction, no usage limit, no minimum order, no stacking rule, no free month and no credit, and no promotion is applied at payment, because the checkout at /checkout-demo is a demonstration in which nothing is charged and no tax is computed. There is no price list, no pocket price waterfall, no discount band report and no margin indicator on the /dashboard. There is no quotation — business/quotes is a page of motivational quotes — and no invoice for the reader's own customers. The volume tiers, retailer codes and affiliate commissions under business/partner-programs belong to La Machine's own distribution program and price La Machine, not the reader's offers. The platform checks no price presentation against the Competition Act or the Consumer Protection Act, as Book B1, Chapter 6, already states; the ordinary-price evidence and the calendar discipline of this chapter are the reader's to keep elsewhere.

Self-Check

1. Why is a ten percent discount on the kit in the running case not a ten percent problem, and what two numbers express its real size?
2. What are the six lines of the pocket price waterfall in this chapter, and which three of them never reach an invoice?
3. Why must a margin floor be expressed as a margin and measured on the pocket price rather than on the invoice price?
4. What must a seller obtain before granting a discount, and what are three counterparts that genuinely justify one?
5. What two tests can make a comparison to an ordinary price defensible in Canada, and where do the figures usually quoted for them actually come from?

Answers. (1) Because the discount comes out of the margin rather than out of the price: $480 off a $4,800 kit is 17.78 percent of the $2,700 margin, and replacing it would take 21.62 percent more volume (Section 6.1.1 and Table 6.1). (2) The list price of $4,800, a volume discount of $288.00, a negotiated discount of $225.60, extended payment terms costing $42.28, a free sensor pack costing $120.00 and an unbilled installation overrun of $150.00, leaving a pocket price of $3,974.12; the last three never reach an invoice and total $312.28 (Section 6.2.2, Table 6.2 and Figure 6.1). (3) Because a discount percentage silently changes meaning whenever a cost moves, while a margin rate recomputes itself; and because the invoice price overstates the margin by every concession granted after it, which on the case is the difference between a 53.46 percent invoice margin rate and a 47.16 percent pocket margin rate (Sections 6.3.1 and 6.3.2). (4) A counterpart, priced in dollars before the reduction is agreed; volume in one order, a multi-year term with indexation, payment in advance or a larger deposit, a reference the business may publish, reduced scope, a faster decision with a signed date, or standard rather than custom terms, of which reduced scope is the safest because it genuinely lowers the seller's cost (Sections 6.5.1 and 6.5.2, Table 6.4). (5) The volume test and the time test of subsections 74.01(2) and 74.01(3), which are disjunctive; the figures of more than fifty percent of sales over twelve months and more than half of a six-month period come from the Competition Bureau's published guidance on ordinary price claims, not from the statute (Section 6.7.2).

Chapter References

Baker, W. L., Marn, M. V., & Zawada, C. C. (2010). The price advantage (2nd ed.). Wiley. https://www.wiley.com/en-us/The+Price+Advantage,+2nd+Edition-p-9781118267653 (publisher page read September 19, 2026; the second edition is by Baker, Marn and Zawada, Baker having replaced Roegner of the first edition)

Competition Act, RSC 1985, c C-34, ss. 45, 52, 74.01, 74.05, 74.1 and 76 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Competition Bureau Canada. (2014, September 15; modified 2024, June 27). Price maintenance (section 76 of the Competition Act): Enforcement guidelines. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/en/how-we-foster-competition/education-and-outreach/price-maintenance-section-76-competition-act

Competition Bureau Canada. (2022, June 23). Ordinary price claims. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/how-we-foster-competition/education-and-outreach/publications/ordinary-price-claims

Consumer Protection Act, CQLR c P-40.1, s. 224 (Québec). LégisQuébec, consolidation of April 7, 2026, as read for Book E1 on September 8, 2026; not read again, LégisQuébec returning an access error on September 19, 2026.

Faustino, S. (2025, May 18). SaaS billing report: Billing is the invisible force behind your growth. ChartMogul. Retrieved September 19, 2026, from https://chartmogul.com/reports/saas-billing-report/ (aggregated billing data from more than 2,500 subscription software companies for full-year 2024; a non-probability sample of one vendor's own customers, excluding companies below $300,000 of annual recurring revenue)

Guide Concerning Use of the Word "Free" and Similar Representations, 16 C.F.R. § 251.1 (2026) (United States). Canonical source https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-251; text read September 19, 2026, at the Legal Information Institute, Cornell Law School, which publishes no consolidation date.

Guides Against Deceptive Pricing, 16 C.F.R. pt. 233 (2026) (United States). Canonical source https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-233; text read September 19, 2026, at the Legal Information Institute, Cornell Law School, which publishes no consolidation date.

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Marn, M. V., & Rosiello, R. L. (1992, September–October). Managing price, gaining profit. Harvard Business Review. https://hbr.org/1992/09/managing-price-gaining-profit (article page and publisher product description read September 19, 2026; the full text was not read, and the pagination reported by secondary sources could not be verified at source)

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Office de la protection du consommateur. (2025, February 10). Annonce d'un prix tout inclus. Gouvernement du Québec. Retrieved September 19, 2026, from https://www.opc.gouv.qc.ca/commercant/pratique-commerce/publicite-loi/prix-tout-inclus

Office de la protection du consommateur. (2026, January 7). Politique d'exactitude des prix. Gouvernement du Québec. Retrieved September 19, 2026, from https://www.opc.gouv.qc.ca/commercant/pratique-commerce/prix-rabais/politique

Robinson-Patman Act, 15 U.S.C. § 13 (2024) (United States). Canonical source https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title15-section13; text read September 19, 2026, at the Legal Information Institute, Cornell Law School.

Sherman Act, 15 U.S.C. § 1 (2018) (United States). Canonical source https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title15-section1; text read September 19, 2026, at the Legal Information Institute, Cornell Law School.

U.S. Department of Justice, Antitrust Division. (n.d.). Price fixing, bid rigging, and market allocation schemes: What they are and what to look for [Antitrust primer]. Retrieved September 19, 2026, from https://www.justice.gov/atr/price-fixing-bid-rigging-and-market-allocation-schemes

Chapter 7 — Quotes and Price Presentation

Groupe immobilier Vallières manages twelve commercial and rental buildings in the Eastern Townships, and its asset manager asked Hydrosentinelle inc. for a price on a Monday afternoon. Karim Haddad, who leads sales and operations, had by then a list price, a volume band, three service tiers, a margin floor and an authority limit, all of which the previous five chapters built. What the company did not have was a document. The first draft showed a single total, no taxes, no options, no deposit, no expiry date, and a sentence saying installation would be scheduled "at a mutually convenient time." That document would have cost the company money in four separate ways, and none of them had anything to do with the price being wrong. This chapter is about the distance between a price that is correct and a price that is presented correctly. It settles whether a price should be published or quoted deal by deal; it builds the price page and the quotation block by block; it explains how the three options are manufactured and where their spacing comes from; it applies the rule that the first price a buyer sees is the price the buyer pays apart from taxes; and it prices the deposit, the payment term and the validity period, which are pricing decisions disguised as administration. It stays, from first line to last, on the side of the document. Presenting that document, defending the number against an objection and trading a concession for a commitment belong to Book E1 (The Sales Process: From Prospect to Signed Contract), Chapter 7, and this chapter does not reopen them.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Decide whether to publish a price, publish a starting point or quote deal by deal, using five conditions you can check against your own offer in an afternoon.
• Build a price page that answers a buyer's five questions in ninety seconds, and that cannot be read as a price the buyer will not actually be charged.
• Assemble a quotation from its ten blocks, and say what each block prevents from going wrong later.
• Manufacture a set of three options with a defensible gap, a real top option and a known margin on each, rather than copying a pattern.
• Display a total price and its taxes lawfully, in Canada and, if you sell south of the border, under the American regime, which is built differently.
• Price the deposit, the net term and the validity period, and say in dollars what each one is worth and what each one costs.

Why It Matters

A price that is never presented is not a price; it is an intention. Everything the previous chapters produced — the value estimate, the tier structure, the floor, the authority limit — reaches the buyer through two surfaces, a published page and a quotation, and those surfaces are where a well-built price is most often lost. The loss is rarely dramatic. It happens because a total was assembled in a way that invited a line-by-line negotiation, because an option was written so vaguely that the buyer chose the cheapest one by default, because no expiry date was given and the document came back six months later at a cost base that had moved, or because a mandatory charge appeared at the end and turned a fair price into a suspicious one.

For a solo founder, the document is the reputation: a quotation is often the first artefact a buyer sees that was produced by the business rather than by the person. For a small team, the document is the policy made visible, because the moment two people send quotations from two templates, the business is quoting two scopes, two payment terms and, in effect, two prices. For an established business, the document is the evidence: every question asked two years later — what was included, what a discount was traded for, whether the ordinary price was real, when the price expired — is answered by the quotation or by nothing at all.

There is a legal dimension here sharper than in any other chapter of this book except Chapter 6. A published price and a quoted price are both representations. In Canada, a price the buyer cannot obtain because of a fixed obligatory charge added later is a false or misleading representation, and selling above an advertised price during the period the advertisement covers is reviewable conduct in its own right. Those provisions are breached most often by small businesses with an out-of-date page and a template nobody owns.

In This Chapter

• 7.1 Publish or Quote: How the Decision Is Made
• 7.2 The Price Page
• 7.3 The Anatomy of a Quote
• 7.4 Options Inside a Quote
• 7.5 Total Price, Taxes and Partial Price Representation
• 7.6 Deposits, Payment Terms and Validity
• 7.7 Multi-Site, Multi-Year and Multi-Currency Quotes
• 7.8 After the Quote: Expiry, Revision and Records

7.1 Publish or Quote: How the Decision Is Made

7.1.1 Two surfaces, two different objects

A price reaches a buyer by one of two routes, and they are not variants of each other. A published price is a standing representation made to everyone who can see it, for as long as it is visible, without knowing who is reading. A quoted price is a representation made to one named buyer, for a defined scope, for a defined period, and it carries the terms on which that exchange is proposed.

The difference has consequences owners discover late. A published price must be maintained: it is reviewable conduct in Canada to advertise a product for sale in a market and then, during the period and in the market the advertisement covers, to supply it at a higher price (Competition Act, s. 74.05). A page that says one number while the quotation template says another is not a marketing inconsistency; it is the breach. A quoted price must instead be bounded, because a quotation carrying no end date creates an obligation the business did not intend and cannot price. Most small businesses never choose. They drift into a hybrid in which a page displays an old number nobody updates while every real transaction is quoted, which is the worst of both, because the published number becomes an anchor the seller then argues against with the business's own buyers.

DEFINITION Quote, proposal, estimate, order form and price page A quote, also called a quotation, is a document stating a firm price for a defined scope, on defined terms, addressed to a named buyer and open for acceptance for a stated period; when the buyer accepts within that period and on those terms, the parties usually have an agreement, which is why the period matters. An estimate is a considered prediction of what work will cost, expressly not a firm price, used when the scope cannot be pinned down before the work starts; the difference between the two words is a difference in who carries the risk of the scope being wrong, and it belongs in the document rather than in the reader's knowledge of the convention. A proposal is a longer document that contains a quote and adds the case for buying at all: the problem restated, the approach, the evidence, the team and the timetable. A proposal persuades; a quote prices. An order form is the acceptance instrument, restating the selected option, the quantities, the total, the taxes and the payment terms, so that what was accepted is unambiguous. A price page is a published, standing statement of prices addressed to no one in particular and to everyone at once. A price book is the internal document holding current prices, costs, discount bands and effective dates, from which pages and quotes are generated. In plain terms: a price book is what you know, a price page is what you say to everyone, a quote is what you say to one buyer, and an order form is what the buyer says back.

7.1.2 The buyer is comparing whether or not you participate

The argument for publishing is not transparency as a virtue. It is that a buyer in the research phase will compare, and a business whose number cannot be found is compared anyway, on the assumption that the missing number is high. Book B1 (Marketing Fundamentals for Business Owners), Chapter 6, states the three readings a price produces, and this chapter develops the third: a price well below the alternatives says "cheaper, and probably less"; a price well above them says "better, and here is why"; and a hidden price says "expensive, and embarrassed."

The strategic pricing literature treats price communication as a discipline of its own rather than as an afterthought to price setting; the seventh edition of Nagle, Müller and Gruyaert devotes a chapter to price and value communication, sequenced between economic value and price structure (Nagle et al., 2023). For a business of this size the practical form is simple. Publishing buys qualification, because buyers who cannot afford the offer disqualify themselves before they consume an hour of the owner's time. Not publishing buys flexibility, which is legitimate when buyers' situations genuinely differ and merely opportunistic when they do not.

KEY CONCEPT A hidden price says "expensive, and embarrassed" The decision not to publish a price is itself a message, and not the message the seller intends. The seller hears "our pricing depends on your situation." The buyer hears that the number is large enough to need a conversation before it can be said, and that the conversation exists to find out how much this particular buyer can be charged. That second reading is what makes a hidden price expensive: it does not merely delay the comparison, it poisons it, because the buyer now suspects the price is a function of the buyer rather than of the offer. Three consequences follow. The business attracts enquiries from buyers who could never have afforded it, and pays for them in hours. It loses buyers who never enquired at all and who therefore never appear in any report the business keeps, which is why this cost is invisible and why owners underestimate it. And it arrives at the first conversation with no anchor of its own, so the first number in the room is whatever the buyer imagined, which is almost always lower than the number the business would have published. The remedy is not always to publish the full price. It is to publish something true and useful — a starting price a real buyer can obtain, a range with the conditions that place a buyer inside it, or the price of a defined first step.

7.1.3 The five conditions that make publishing safe

Publishing is right when five conditions hold together, and quoting is right when any one fails. First, the offer is standard enough that the same number is correct for most buyers who will read it. Second, the cost to serve does not vary by more than the margin can absorb; a business whose delivery cost doubles between an easy site and a difficult one cannot publish one installed price without either losing money on the difficult sites or overcharging the easy ones. Third, no negotiation is expected, or the published price is explicitly a list price from which stated, published reductions apply. Fourth, the business can honor the number everywhere it is visible, for the whole period it is visible. Fifth, the business can keep the page current, with an effective date and a change record, because an out-of-date price page is the commonest way a small business supplies above its advertised price without ever deciding to.

Hydrosentinelle satisfies the first three for its subscription and fails the second for its kit. Monitoring is the same service in every building at a given tier, its cost to serve is stable and its price is not negotiated, so the company can publish $89, $129 and $179 a month with the contents of each tier beside them. The installed kit is different: the number of sensors, the difficulty of the runs, the state of the mechanical room and the coordination with an existing building system all move the installation cost, and the volume band moves the price. For the kit, the company publishes a starting point and quotes the rest.

The table below sets the three postures against the conditions that select them. Its columns give the posture, when it is right, what it buys, what it costs and the case's own choice, so a reader can place each of their own offers in one of the three rows rather than adopting one policy for everything they sell. Read the fourth column hardest: every posture has a real cost, and the mistake is believing one of them is free.

Table 7.1 — Publish, publish a starting point, or quote

7.1.4 The honest starting point, and where it becomes a partial price

A starting price is the most useful of the three postures for a small business and the easiest to get wrong. It is honest when a real buyer can obtain the stated configuration at the stated price, when the conditions that place a buyer in that configuration sit beside it rather than in a note at the bottom, and when the things most buyers will additionally need are named with their prices or with the way their price is determined. It becomes a partial price representation the moment the stated number cannot be attained because of a charge the buyer must pay and the business has not disclosed up front. Chapter 3 draws that line for framing; Section 7.5 applies it to a document.

Two variants deserve mention. A published range works only when the conditions that place a buyer at each end are stated. A configurator, on which the buyer assembles the offer and watches the total change, is the strongest form of publishing, because the total is one the buyer built; Chapter 12 examines what a tool must do to support one. Both obey the same condition as any published price: the number the buyer arrives at must be the number the buyer will be charged, apart from taxes.

7.2 The Price Page

7.2.1 What a page has to accomplish in ninety seconds

A price page is read quickly, once, by someone deciding whether to continue. It must answer five questions in that time: who this is for, what is included, what it costs, what is not included and what happens next. A page that answers four of the five sends the reader to a competitor to answer the fifth. The commonest omission is the fourth, because pages describe inclusions at length and leave exclusions to be discovered in a conversation, which is exactly the discovery that destroys trust in the number.

The second job is arithmetic. A buyer comparing two offers converts both to a common unit — a year, a building, a seat — and a page that does not perform that conversion leaves it to someone with an incentive to be pessimistic. Hydrosentinelle's page shows the monthly tier price, the annual equivalent and the price per building. The third job is to remove the fear of the unknown total, because a buyer who cannot compute one assumes the largest plausible figure; a worked total for a typical buyer, labeled as an example, advances a decision more than any amount of feature description.

7.2.2 The eight blocks of a price page

A page that does all three jobs contains eight blocks, and their presence matters more than their order. The table below names each block, states what it must contain to be useful, and names the failure that follows when it is missing. Its columns run from the block to the failure, and the last column is the one to read first, because most readers will recognize a failure they have experienced before they recognize a block they have never written.

Table 7.2 — The eight blocks of a price page, and what each one prevents

7.2.3 When a published page becomes an offer

An owner's mental model of a price page is usually that it is advertising, and that nothing binds until a contract is signed. That model is close enough in ordinary commercial dealings and wrong in one case that matters in Québec. A merchant whose published proposal contains all the essential elements of a distance contract is deemed to make an offer, even where the document states the contrary, and the consumer statute is public order, so a clause disclaiming the effect does not remove it (Consumer Protection Act, ss. 54.1, 261 and 262, as read for Book H1 on September 13, 2026, at the consolidation of June 12, 2026). A business selling to consumers online should therefore treat its price page as a document with legal effect rather than as a brochure.

The federal position points the same way. The exceptions to section 74.05 are narrow — a catalog prominently stating that prices may be in error, where the advertiser proves the error; an advertisement immediately followed by a correction; securities; and supply by a person not in the business of dealing in the product — and the market covered is the one the advertisement could reasonably reach, unless it expressly restricts itself. For a small business the consequence is one sentence: if the page says it, the quotation honors it until the page changes.

WHAT CHANGED Minimum advertised prices moved to the reviewable track in 2009 A supplier that sells through resellers frequently wants to control how its price is displayed, and the rules changed within the working lives of most owners. Price maintenance was a criminal offense in Canada until 2009; since then it has been a reviewable practice under section 76 of the Competition Act, which means an order now requires two findings rather than one. The Tribunal must find, first, that the person influenced upward or discouraged the reduction of the price at which a customer or reseller supplies, offers or advertises a product, by agreement, threat, promise or any like means, or refused to supply or otherwise discriminated against someone because of that person's low pricing policy; and, second, that the conduct has had, is having or is likely to have an adverse effect on competition in a market. Two subsections speak directly to presentation. Suggesting a resale price is itself evidence of upward influence unless the supplier has made clear that the recipient is under no obligation to accept it and that business relations will not suffer in any way if it is not accepted; and a non-retailer advertising a resale price is presumed to be exerting influence unless the advertisement makes clear the product may be sold at a lower price. The Competition Bureau's enforcement guidelines, published September 15, 2014 and modified June 27, 2024, add that the adverse-effect threshold sits below the substantial-lessening standard of other provisions, that without market power there is generally no adverse effect, and that a party directly affected may seek leave to bring its own application (Competition Bureau Canada, 2014). What changed is that the question is no longer whether an understanding about a displayed price existed but whether it hurt competition. Chapter 6 treats the policy side of the same provision. Verify the current text before relying on it.

7.2.4 What a page must never do

Three things destroy a price page, and all three are done with good intentions. The first is a mandatory charge disclosed below the fold: an administration fee, a mandatory setup, a compulsory first-year service. The second is a comparison price the business cannot document; Chapter 6 sets out the ordinary-price tests, and a page is the easiest place to breach them because nobody remembers when the crossed-out number was last charged. The third is an option presented as optional that the offer does not work without. The remedy for all three is one question, asked of every number before it is published: can a real buyer pay this amount and receive a working offer?

7.3 The Anatomy of a Quote

7.3.1 Ten blocks, and what each one prevents

A quotation is a small contract proposal wearing the clothes of a price list, and it is built from ten blocks. Each block exists because its absence produces a specific, repeatable loss. The table below names the ten, states what each must contain, and names the loss that follows when the block is omitted. Its columns run from the block to the consequence; readers who already send quotations should read the third column first and count how many of those losses they have paid for in the last year.

Table 7.3 — The ten blocks of a quotation, and the loss each one prevents

7.3.2 The price lines, and how far to break them down

A real decision hides in the fifth block. Breaking a price into many lines makes the document transparent and every line negotiable; a single total makes the offer whole and makes the buyer negotiate the whole. The workable rule is to break the price along the lines the buyer will actually vary — quantity, sites, tiers, optional items — and to keep whole everything the buyer cannot buy separately. Hydrosentinelle quotes twelve kits as twelve units at one unit price, because the buyer can vary the count; it does not itemize sensors, gateway, labor and commissioning inside the kit, because an itemized kit invites a discussion about the price of a sensor, which is a discussion about cost rather than about value. The same rule decides what appears as an add-on: an add-on is a line the buyer may genuinely decline and still receive a working offer.

7.3.3 The assumptions block, which most quotations omit

Between the scope and the price sits a block small businesses almost never write and that prevents more disputes than any other: the assumptions on which the price depends. For an installed product these are the state of the site, the availability of access, the presence of a functioning network, the completion of work by others, and the schedule. The block is short and specific. It says what the business assumed, and what happens if the assumption is wrong: a stated hourly rate, a change order, or a re-quote. Without it, the business discovers the variance during delivery and absorbs it, which Chapter 6 measures on one deal as an unbilled installation overrun that never appears as a discount anywhere in the records but reduces the pocket price exactly as a discount would.

METHOD Building a quote in nine steps When to use it: every time a price is proposed to a named buyer for a defined scope, and again whenever a quotation template is rebuilt after a price change. Steps: (1) Write the scope before any number, listing deliverables, quantities, sites, the timetable and the point at which the work is accepted, and list beside it what is expressly excluded. (2) Write the assumptions the price depends on, and state for each what happens if it proves wrong. (3) Price the base configuration from the price book, at list, applying only the published bands the buyer's volume entitles them to, and record which band was applied. (4) Build the options as complete, buyable configurations rather than as add-on menus, name the recommended one, and check that each option's margin is above the floor. (5) Total the pre-tax lines, then apply each tax on its own line, computed on the stated base, so the total payable appears once and unambiguously. (6) Set the payment terms: deposit as an amount and a percentage, the net period, any milestones, and the currency of every figure. (7) Set the expiry date, and state what happens after it, which is normally that the quotation must be reissued at current prices. (8) Have it reviewed against the floor and the authority limits by someone who did not negotiate it, and record any approval given. (9) Send it as one document with a version number, and log the send with its date, total and expiry. Deliverable: a quotation of one document containing the ten blocks of Table 7.3, with the recommended option named and every figure traceable to the price book. Verification criterion: a colleague who was not involved can read it and state, without asking a question, what is included, what it costs in total, what tax is added, what is due on acceptance and the last day it may be accepted.

7.3.4 The quotation of the case, in five blocks

Figure 7.1 draws the Vallières quotation as the five blocks the buyer actually reads, in the order the buyer reads them: the hardware and its total, the taxes added to it, the deposit due on acceptance, the three monthly options, and the conditions that bound the document. At the foot it prints the first-year total before taxes under each option, which is the arithmetic the buyer would otherwise do badly. Sections 7.4 to 7.6 build each of the five blocks in turn.

Figure 7.1 — The Vallières quotation of the case, in five blocks. The figure sets out the structure of a quotation for twelve buildings addressed to Groupe immobilier Vallières by Hydrosentinelle inc. The first block is the hardware: twelve installed kits at $4,512.00 each, after the published volume band of 6 percent that applies from five buildings, for $54,144.00 before taxes. The second block is the taxes, each on its own line and each computed on that pre-tax base: the Goods and Services Tax at 5 percent is $2,707.20 and the Québec Sales Tax at 9.975 percent is $5,400.86, for a total payable of $62,252.06. The third block is the deposit, 30 percent of the pre-tax hardware total, or $16,243.20, due on acceptance, leaving a balance of $37,900.80. The fourth block is the monthly monitoring, quoted as three complete options for the twelve buildings: Essential at $1,068.00 a month, Plus at $1,548.00 a month and Portfolio at $2,148.00 a month, the difference between the top two being $600.00 a month. The fifth block is the conditions: the quotation is valid for 30 days and the balance is payable net 30 days. At the foot, the first year before taxes under each option: $66,960.00 with Essential, $72,720.00 with Plus and $79,920.00 with Portfolio. Hydrosentinelle inc. is fictional; every figure is computed by one model.

WHO DOES THIS Who signs a quote, and who may not In a solo business, the owner prepares, approves and signs, and the only protection against the owner's own end-of-month arithmetic is that the floor and the authority limit were written on a different day, in writing, and are treated as binding; the practical device is a two-minute checklist at the foot of the template. In a small team, the person who prepares a quotation and the person who approves a departure from list should not be the same person, and the signatory should be the person accountable for the margin rather than the one whose income depends on the deal closing. At Hydrosentinelle, quotations are prepared by Karim Haddad, who leads sales and operations, within the authority limit Chapter 6 sets at 6 percent for a representative and 12 percent for the head of sales; anything below the floor requires Laurence Bergeron, the president, to approve it in writing before the document is sent, not afterwards. In an established business, three further roles appear: someone owns the template and the price book, someone reviews non-standard terms before they are offered rather than after they are accepted, and someone records every quotation sent so that win rate and discount can be measured. On the buyer's side, a quotation should name the capacity in which it is accepted. Book H1 (Contracts and Commercial Agreements) treats signing authority and the formation of the agreement; this chapter treats only what the document must say.

7.4 Options Inside a Quote

7.4.1 What this book builds, and what Book E1 presents

Three options in a quotation are among the most reliable devices in commercial practice, and the division of labor in this library is precise. Book E1, Chapter 7, owns the presentation: how the options are introduced in a conversation, why the anchor is set on the cost of the problem rather than on the price of the product, and why the recommended middle option is chosen most often. That chapter relies, as this one does, on the finding that a judgment about a number is pulled toward whatever number was considered first, among the most replicated results in decision research (Kahneman, 2011). This chapter owns the manufacture: what each option contains, how far apart they are, what margin each carries, why the top one must be real, and where the level of the anchor comes from. Chapter 3 explains the mechanism, and Chapter 5 builds the tier structure from which the options are drawn.

The level of the anchor is the part most often improvised, and it should not be. It comes from the value estimate of Chapter 2, not from the seller's comfort and not from a competitor's list. For the case, the economic value of avoiding water damage in a building was estimated and the company's price captures 33.66 percent of it, so the top option is anchored on a share of a value the buyer can recognize, and the document can say so in one sentence without becoming a sales conversation. An anchor the business cannot justify from the buyer's own economics is only a number the seller hopes will work.

7.4.2 How far apart the options should be

The gap between options does three things at once, which is why it cannot be set by a rule of thumb. It must be large enough to matter, so that choosing the higher option is a decision rather than a rounding error. It must be small enough to climb, so that a buyer who wants the higher option can justify the step to whoever approves budgets. And it must be explainable by contents, so the difference in price corresponds to a difference the buyer can name. When a buyer cannot name what the extra amount buys, the gap reads as a margin grab, and Chapter 5 makes the same point structurally: a tier a buyer cannot tell apart from the next one is a discount waiting to be requested.

On the Vallières quotation the three options are the three tiers of Chapter 5, multiplied by twelve buildings. Essential is $1,068.00 a month, Plus is $1,548.00 a month and Portfolio is $2,148.00 a month. The step from Essential to Plus is $480.00 a month and the step from Plus to Portfolio is $600.00 a month; over a first year those steps are $5,760.00 and $7,200.00. Measured against the first-year total, the step from Plus to Portfolio is 9.90 percent of a $72,720.00 commitment, which a building portfolio manager can take without a new approval, and that is precisely why it sits where it does.

The table below gives the three options as the quotation presents them, with the monthly amount for twelve buildings, the monthly gross margin each carries at the tier costs of Chapter 5, the first-year total before taxes including hardware, and the step from the option below. The margin column is there for the seller rather than the buyer: it is what makes an honest recommendation possible, because the business knows what each option is worth to it.

Table 7.4 — The three options of the Vallières quotation, for twelve buildings

7.4.3 The top option has to be one somebody buys

A top option that exists only to make the middle one look reasonable is a decoy, and a decoy nobody could ever want is two problems rather than one. Commercially it fails, because buyers detect an option that has been constructed rather than offered, and the detection discredits the whole document. Legally it is a representation like any other, judged on the general impression it conveys, and a representation about an offer the business is not prepared to deliver falls on the wrong side of the line Chapter 3 draws. The test is operational and takes five minutes. Can the business deliver the top option tomorrow, at the price shown, to the buyer receiving the quotation? Does a named customer already have it, or is there a named prospect for whom it is plainly the right answer? Is its cost to serve known, so that its margin is known? Hydrosentinelle passes all three for Portfolio, which requires a minimum of five buildings and which Vallières, with twelve, comfortably exceeds; the fence is a real fence rather than a label, and it is published in the tier description rather than invented for this quotation.

7.4.4 An option nobody chooses is mis-built

Book E1, Chapter 14, sends readers to this book with two named symptoms, and one of them is options nobody chooses. The diagnosis belongs here because the cause is in the construction rather than the presentation. There are four causes and they are distinguishable. The option may be undifferentiated, so the buyer cannot name what the extra amount buys; the remedy is in the contents, which Book A6 (Designing a Value-Added Offer) designs. It may be mis-spaced, the step being too large to climb or too small to be worth taking. It may be mis-recommended, the document steering toward the wrong option for the buyer in front of it. Or it may be unbuyable, blocked by a fence the buyer cannot pass, such as a minimum the buyer does not reach.

Two of the four are cured in the document and two are not. Chapter 9 supplies the measurement that tells them apart: the distribution of accepted options over a year, read beside the distribution of quoted options. An option quoted often and accepted never is mis-spaced or undifferentiated; an option rarely quoted at all is blocked by a fence or by the person preparing the document.

7.5 Total Price, Taxes and Partial Price Representation

7.5.1 The principle both countries share

Before any provision is cited, one principle is now expressed by Canadian and American law in their own vocabularies, and a business that holds it will rarely be surprised by either. The first price a buyer is shown must be a price the buyer can actually obtain, and the only amounts that may be added afterwards are amounts imposed by government. Everything else follows: mandatory fees belong inside the first number, optional items may be shown separately because they are genuinely optional, and the total payable must be at least as prominent as any component of it. Book B1, Chapter 6, states the practical version for this library — the price the buyer sees first is the price they pay, apart from taxes, and every crossed-out price is real and documented — and this chapter demonstrates it provision by provision rather than restating it.

The principle applies to a quotation as much as to a page, and the quotation is where small businesses breach it most often, because it is assembled by hand. A commissioning fee added below the total, a shipping line the buyer cannot avoid, a compulsory first-year service presented as an option, and a minimum that appears only in the terms are the four forms it takes.

THE RULE Competition Act, s. 74.01(1.1): the first price shown must be one the buyer can actually pay Jurisdiction: Canada, federal. Instrument: Competition Act, RSC 1985, c C-34, s. 74.01, in the consolidation current to July 21, 2026 and last amended March 26, 2026, read on the Justice Laws Website on September 19, 2026. Subsection 74.01(1.1) provides that the making of a representation of a price that is not attainable because of fixed obligatory charges or fees is a representation that is false or misleading in a material respect, except where those charges or fees represent only an amount imposed by or under an Act of Parliament or of the legislature of a province. Two features govern its use. First, it sits in Part VII.1, which is civil: the conduct is reviewable, no intention has to be proved, and it is not necessary to show that anyone was deceived; subsection 52(1.3) is the criminal mirror and adds a knowing or reckless state of mind. Second, the exception is narrow: only amounts imposed by legislation qualify, which in ordinary commerce means sales taxes and equivalent statutory charges, and not an administration fee, a mandatory setup, a documentation charge or a compulsory service, however customary. Penalties are set by section 74.1: for a corporation, a first order may carry the greater of $10,000,000 and three times the benefit derived, or, where that benefit cannot reasonably be determined, three percent of annual worldwide gross revenues. Scope for a reader of this book: every price page, advertisement, quotation, order form and invoice the business issues, including those a system displays on its behalf. Verify the current text before relying on it.

7.5.2 Where the taxes go in a quotation, and why they go there

Taxes are the one set of amounts a business may lawfully add to a displayed price, and the way they are added is itself a presentation decision. The rule for a document is straightforward: each tax appears on its own line, the base on which it is computed is visible, and the total payable appears once, after them, more prominently than any component. Book E1, Chapter 7, makes the same requirement from the contract side, listing taxes stated separately among the financial clauses a proposal must carry, and this chapter supplies the arithmetic.

On the Vallières quotation the pre-tax hardware total is $54,144.00. The case model computes both Québec taxes on that same pre-tax base: the Goods and Services Tax at 5 percent is $2,707.20 and the Québec Sales Tax at 9.975 percent is $5,400.86, so the taxes together add 14.975 percent, or $8,108.06, for a total payable of $62,252.06. Registration, collection, input tax credits and remittance are the subject of Book G5; this chapter takes only the display decision, and the two taxes are separate lines because they are separate statutory charges and the buyer must be able to reconcile the document to an invoice.

One nuance decides which regime governs the document, and it is frequently missed. Groupe immobilier Vallières is a business buying for its own commercial operations, so the Québec consumer statute does not govern this quotation; the Competition Act, which is not limited to consumers, does. A reader who sells the same product to a person for personal use is governed by both at once, and the consumer regime is the stricter of the two on display.

QUÉBEC & CANADA NOTE GST and QST in a Québec quotation, and the order of the lines In Québec, the advertised price must correspond to the total amount the customer has to pay. The Office de la protection du consommateur, which enforces the Consumer Protection Act, states on its page of February 10, 2025 that administration fees, transport and preparation fees, and charges arising from legal obligations that are not remitted to a public authority must all be inside the advertised price, while the Goods and Services Tax, the Québec Sales Tax and specific contributions remitted to a public authority may be added at payment; the same page adds that the total price must be displayed more prominently than its components, and that for a package the global price predominates over the price of each element (Office de la protection du consommateur, 2025). The statutory basis for the display obligation is section 224 of the Consumer Protection Act, under which a merchant may not, in an advertisement, give less prominence to the total price of an assembly of goods or services than to the price of a part of it, may not disclose the amount of installments without disclosing and giving prominence to the total price, and may not charge a higher price than the price advertised; that is the reading made at source for Book E1 on September 8, 2026, at the consolidation of April 7, 2026, and LégisQuébec returned an access error on September 19, 2026, so the text was not read again for this book. Québec also requires merchants to indicate the price on products, with categories of exemption and a compensation policy for scanning errors; that requirement is stated here on the authority of the Office as the enforcing body, because the text of the provision could not be read, and this book does not quote a provision it has not read. The practical consequence for a quotation is an order of lines that never varies: the priced scope, the pre-tax subtotal, each tax on its own line, and then the total payable, set apart and more prominent than anything above it. For the case that order gives $54,144.00, then $2,707.20, then $5,400.86, then $62,252.06.

7.5.3 The same question answered in the United States

A reader who quotes American buyers meets a regime built on different foundations, through two instruments of unequal weight. The Guides Against Deceptive Pricing are interpretive guides issued under the general prohibition on unfair or deceptive practices; they state what the Federal Trade Commission considers deceptive and create no free-standing civil penalty (16 C.F.R. pt. 233). The Rule on Unfair or Deceptive Fees is a trade regulation rule, effective May 12, 2025, carrying civil penalties, and it is narrow, covering tickets to live events and short-term lodging rather than the economy at large (16 C.F.R. pt. 464). What matters for a document is its definition of the total price as the maximum total of all fees a consumer must pay for the goods or services and any mandatory ancillary good or service, excluding only government charges, shipping charges and genuinely optional items, and shipping qualifies only where it reasonably reflects what the business incurs. The Commission's published questions and answers add that the rule protects individual and business purchasers alike, applies online, in an application and in a store, and prohibits no type or amount of fee, no itemization and no dynamic pricing (Federal Trade Commission, 2025).

The table below compares the two regimes on the five questions a document actually raises. Its columns give the question, the Canadian answer, the American answer and the instrument in each country, so a reader quoting on both sides of the border can check a template once rather than guessing twice. Read the third row most carefully: the instruments differ in kind, not merely in wording, and treating an interpretive guide as a penalty-bearing rule, or the reverse, is the error this comparison exists to prevent.

Table 7.5 — How Canada and the United States treat the same five display questions

DO THE MATH The Vallières quote for twelve buildings, line by line, with taxes Fictional example; figures are invented and computed by the case model. Groupe immobilier Vallières asks Hydrosentinelle inc. to price twelve buildings. The list price of an installed kit is $4,800.00; the published volume band that applies from five buildings is 6 percent, so the unit price on this quotation is $4,512.00 and twelve units come to $54,144.00 before taxes. The band is worth $3,456.00 against a list total of $57,600.00, and it is applied because the quantity entitles the buyer to it, not because it was requested. The Goods and Services Tax at 5 percent on that base is $2,707.20 and the Québec Sales Tax at 9.975 percent is $5,400.86, so the taxes add $8,108.06 and the total payable is $62,252.06. The deposit is 30 percent of the pre-tax hardware total, or $16,243.20, which is 26.09 percent of the total payable and covers 64.46 percent of the $25,200.00 the twelve kits cost the company to build and install; the balance of $37,900.80 is payable net 30 days. Monitoring is quoted as three complete options for twelve buildings: Essential at $1,068.00 a month, Plus at $1,548.00 a month and Portfolio at $2,148.00 a month, which is $600.00 a month above Plus. Adding twelve months of monitoring to the hardware gives a first year before taxes of $66,960.00, $72,720.00 or $79,920.00 according to the option chosen. The quotation is valid for 30 days.

7.6 Deposits, Payment Terms and Validity

7.6.1 A deposit is a pricing decision wearing an administrative disguise

Book B1, Chapter 6, names payment terms as a signal and leaves the arithmetic to this book: a deposit signals seriousness on both sides, monthly billing lowers the entry barrier, and a discount for annual payment trades margin for cash and commitment. A quotation must settle the first of the three. A deposit does four things at once, and a business that asks for one without knowing which of the four it wants will set the percentage by imitation.

It finances the hardware: the twelve kits of the Vallières quotation cost Hydrosentinelle $25,200.00 to build and install, and the 30 percent deposit of $16,243.20 covers 64.46 percent of that outlay before the company spends it. It qualifies the buyer, because a buyer who will not pay a deposit is frequently one who has not obtained the internal approval the seller was told existed. It prices the risk of cancellation. And it signals symmetry, since a business that asks for nothing communicates that its own commitment is worth nothing. Book G2 manages the cash gap that remains; this chapter decides the deposit itself, because it is part of the price.

7.6.2 The net period costs money, and the amount is computable

A payment period is a loan the seller makes to the buyer at a rate of zero, and its cost can be computed exactly. The case model carries a cost of money of 12 percent a year, and that figure prices every extension. Extending the Vallières balance of $37,900.80 from net 30 to net 60 costs $373.82 in financing; extending the whole pre-tax total of $54,144.00 by the same thirty days costs $534.02. Chapter 6 measures the same effect on one kit, where thirty extra days on a negotiated price of $4,286.40 costs $42.28 and disappears into the pocket price without ever being recorded as a discount.

Those amounts look small until they are read as what they are: a concession granted at the end of a negotiation, by someone who did not know its price, on every deal. The remedy is not to refuse longer terms but to price them and trade them. A quotation offering net 30 as standard and net 60 at a stated addition has turned an invisible leak into a line the buyer can choose, and Chapter 6's rule of exchange then applies to it like any other concession.

The table below gives the four payment instruments a quotation can use, what each one buys the seller, what each one costs and what the case's quotation does with it. The cost column is the one most businesses have never computed for their own numbers.

Table 7.6 — The four payment instruments of a quotation

7.6.3 The validity period is the cheapest protection in the document

An expiry date costs nothing to write and protects four different things. It protects against a cost base that moves, which for a hardware business means components and labor. It protects capacity, because a quotation accepted five months late lands in a month that is already full. It protects the price structure, because a quotation accepted after a price change becomes a grandfathered price nobody decided to grant; Chapter 8 treats grandfathering as a deliberate instrument, and an old quotation is the accidental version of it. And it protects the negotiation, because a buyer who knows the document expires has a reason to decide.

Thirty days is the case's choice and a common one. What matters more than the number is that the document says what happens afterwards, and the honest sentence is that the quotation must be reissued at the prices in force on the day it is reissued. A business that quietly honors expired quotations has taught its buyers that the date is decorative.

WATCH OUT A quote with no expiry date is an offer you have left open A quotation without an expiry date does not become void after a decent interval. It sits with the buyer and comes back at the worst moment: after a supplier increase, after the annual price review, in a month when the installation crew is already committed, or after the buyer has used it to obtain a better number from a competitor. Three specific costs follow. The first is margin: the case's own price review in Chapter 8 moves the monitoring price, and every quotation still open on that day is a commitment at the old number, granted to buyers who did nothing to earn it. The second is the loss of the deadline as a reason to decide, which is the only form of urgency a document can honestly carry. The third is the awkward refusal: a business that has to tell a buyer that a four-month-old quotation is no longer valid, having never said so in writing, pays for the omission in goodwill at the exact moment it is trying to win the deal. One related omission belongs in the same warning: a quotation whose validity date is present but whose price page still shows the old number has protected nothing, because the page is the representation the buyer will rely on.

CASE IN POINT Twelve buildings, three options and one deposit Fictional example; names and figures are invented. The asset manager of Groupe immobilier Vallières asked Hydrosentinelle inc. for a price on Monday and expected a number by Friday. Karim Haddad's first draft showed one total for twelve buildings, no tax lines, no options, no deposit and no expiry date. Laurence Bergeron, the president, refused to send it, and the two cofounders rebuilt it against the ten blocks of Table 7.3. The rebuilt document prices twelve installed kits at $4,512.00 each, the published volume band of 6 percent being applied because twelve buildings fall in the band that starts at five, for $54,144.00 before taxes; it shows the Goods and Services Tax at $2,707.20 and the Québec Sales Tax at $5,400.86 on separate lines, for a total payable of $62,252.06; it asks $16,243.20 on acceptance, which is 30 percent of the pre-tax hardware, with the balance of $37,900.80 net 30 days; it offers monitoring as three complete options for the twelve buildings, at $1,068.00, $1,548.00 and $2,148.00 a month, with Plus named as the recommended option; and it expires in 30 days. It also carries an assumptions block the first draft did not: network availability at each building, access during business hours, and mechanical rooms in the state described at the site visit, with a stated consequence if any proves wrong. Nothing in the rebuilt document changed the price. What changed is that the buyer could see the total, what was excluded, what was due when, and until when the numbers held.

7.7 Multi-Site, Multi-Year and Multi-Currency Quotes

7.7.1 Multi-site: the band is a published entitlement, not a negotiation

A quotation for several sites raises a question a single-site quotation does not: whether the buyer gets the volume price for a commitment or for an aspiration. Hydrosentinelle's published bands, built in Chapter 5, give no reduction from one to four buildings, 6 percent from five to nineteen, and 12 percent from twenty upward. Vallières receives the 6 percent band because it is buying twelve, and the quotation says so in those words. That matters: a reduction presented as a published entitlement is not a negotiating concession, and it does not teach the buyer that asking produces reductions.

The difficulty arrives when the buyer wants the band price for a phased rollout: four buildings now, eight over two years, at the twelve-building price today. Three answers are honest, and improvising a fourth is how businesses give away a band. Price what is ordered today at the band the quantity earns. Price at the higher band against a committed quantity, written, with a stated consequence if it is not reached, which is a true-up clause Book H1 drafts. Or offer a retroactive adjustment, each phase priced at the band its cumulative quantity has reached, the difference credited when the threshold is crossed. What is not defensible is granting the twelve-building price for a four-building order.

7.7.2 Multi-year: an unindexed price is a discount nobody negotiated

Book E1, Chapter 7, states the point in one sentence this chapter has to cost out: an unindexed multi-year price is a discount nobody negotiated. The arithmetic is unforgiving because it compounds. The Plus option on the Vallières quotation is $1,548.00 a month, which over a thirty-six-month term is $55,728.00 of monitoring revenue at a price fixed on the day the document was signed, while every cost inside it moves.

A quotation proposing a term of more than one year must therefore say one of three things, and saying nothing is not one of them. It can state a fixed percentage applying on each anniversary. It can name a public index and the month whose value is used, Statistics Canada's published series being the usual source in Canada. Or it can state a cap, the compromise most buyers accept: the price moves by the index, to a maximum of a stated percentage. Chapter 8 sizes the movement; Book H1, Chapters 3, 7 and 9, drafts the clause. This chapter's contribution is the one most often forgotten: the choice must appear in the quotation, not only in the contract that follows, because the quotation is what the buyer compares and approves.

7.7.3 Multi-currency: name the currency on every line

A quotation that crosses a border must state its currency on every figure, not once in a footer, and must say who carries the movement between the date of the quotation and the date of payment. Quoting in the seller's currency moves the risk to the buyer, which is clean and sometimes loses the deal. Quoting in the buyer's currency moves it to the seller, which is a pricing decision and should be priced rather than absorbed. Quoting in the buyer's currency with a stated rate and a stated revision threshold splits the risk, and it is the form a small business can manage. The validity period does real work here, because a thirty-day quotation bounds the exposure to thirty days. The case's own quotation is in Canadian dollars throughout, and it says so on every line.

7.8 After the Quote: Expiry, Revision and Records

7.8.1 Expiry is an event, not an absence

A quotation that reaches its expiry date should produce an action rather than silence. The business notes that the document has expired and, where the opportunity is still alive, reissues it at current prices with a new number and a new date. Doing that consistently has three compounding effects. It keeps the price book and the documents in circulation in step, so a price change actually reaches the market. It makes the expiry date credible. And it produces the record Chapter 9 needs, because a quotation that expires unaccepted is a data point about the price, and a business that does not record expiries measures its win rate against a denominator it has invented.

7.8.2 Revising without teaching the buyer to wait

Every quotation will sometimes be revised, and the way a business revises teaches the buyer what waiting is worth. A revision that lowers the price for no reason other than time has taught the buyer that the first number was not real, and the lesson is applied to every subsequent quotation. A revision that changes the price because the scope, the quantity or a term changed is a different act, and the document should make the reason visible, with a version number, so the buyer sees an exchange rather than a retreat. That is Chapter 6's rule of exchange operating inside a document rather than a conversation. Book E1, Chapter 8, conducts the negotiation that produces the trade; this chapter insists only that the trade be written down.

7.8.3 The records a quotation leaves behind

A quotation generates four records, each consumed by a different chapter of this book. The document itself, with its version, is the evidence of what was offered and on what terms. The log of sends, with the date, the buyer, the total, the option recommended and the expiry, is what Chapter 9 turns into a win rate, a discount distribution and an option-mix reading. The approval trail, recording who authorized any departure from list and against what exchange, makes Chapter 6's authority limits real rather than aspirational. And the price evidence — what the ordinary price was and when it was charged — is what makes a comparison claim defensible under the ordinary-price tests, and it must be retained deliberately because no accounting system produces it as a by-product. None of the four is technically demanding; all four are routinely absent in businesses of this size, and their absence is invisible until the year a question is asked.

TOOLS OF THE TRADE What a quoting tool must do, function by function A quoting tool is judged on nine functions, and a spreadsheet plus a word processor performs several of them badly rather than none at all. Draw from a price book: the document is filled from current prices and costs held in one place, with an effective date. Configure: the preparer assembles a scope from defined items with quantities, and the tool computes the extensions. Apply bands automatically: published volume or term bands are applied by rule from the quantity, and the band applied is recorded. Enforce a floor and an authority: a line below the margin floor cannot be sent without a recorded approval by someone entitled to give it, which is the difference between a policy and a wish. Present options: two or three complete configurations can be quoted side by side, each with its own total and its own recorded margin. Compute taxes correctly: each tax on its own line, on a stated base, with the total payable set apart. Carry terms and an expiry: deposit, net period, milestones, currency and validity are fields rather than free text, so they cannot be omitted. Version and log: every send, revision and expiry is recorded with its date, its total and its author. Export: the log can be extracted in an open format, because Chapter 9 measures from it. Chapter 12 maps these functions against what a spreadsheet, a market tool and La Machine each cover.

IN LA MACHINE There is no quote object, and business/quotes is a page of motivational quotes The route a reader will look for first does not do what its name suggests. Business → Motivational Quotes (business/quotes) is a page of twenty-six hard-coded motivational quotations with a first-name personalization field and a set of categories; it makes no server call for any commercial document, and there is no quotation object anywhere in the platform. The route inventory confirms the absence from the other direction: among the 211 routes there is no quotation, no proposal, no order, no cart, no invoice and no price list (La Machine, 2026). What exists is a published page for a priced item and a client-side purchase path, client/packages and client/store displaying a package or a store product to a customer at its recorded price. What is not here: no expiry date, no validity period and no electronic signature; no option set, no deposit field, no payment term and no net period. Taxes are not computed at payment, /checkout-demo showing a "DEMO MODE" banner, announcing taxes at a next step that never computes them, and debiting nothing, while /client-subscribe is disabled. Routes verified in the route inventory on September 19, 2026, and behaviors read in the application's source at its commit of September 7, 2026.

Costly Mistakes

• Hiding the price and calling it flexibility. A buyer who cannot find a number concludes that it is large and depends on who is asking, and the enquiries the business never received are the cost it never sees.
• Letting a mandatory charge appear after the total. A displayed price that cannot be obtained because of a fixed obligatory charge is a false or misleading representation in Canada, and sales tax is the only addition the exception covers.
• Sending a quotation with no expiry date. The document does not lapse; it returns after the annual price review, at a cost base that has moved, and the business either honors a price it would no longer offer or refuses in writing what it never wrote down.
• Presenting one total instead of options. A single number produces a yes-or-no decision, where three complete, buyable configurations produce a choice that Chapter 9 can measure.
• Granting a payment term without pricing it. Thirty extra days on the balance of the case's quotation costs $373.82 at the model's cost of money, granted by someone who did not know its price and recorded nowhere.
• Quoting a multi-year price with no indexation and no cap. Thirty-six months of the recommended option is $55,728.00 of monitoring at a price fixed on the day of signature, and every cost inside it moves.

Key Takeaways

• A published price and a quoted price are different objects: the published price must be honored everywhere it is visible for as long as it is visible, and the quoted price must be bounded by a named buyer, a defined scope and an expiry date.
• Publishing is right when five conditions hold together — a standard scope, a stable cost to serve, no expected negotiation, the ability to honor the number and the ability to keep the page current.
• A hidden price says "expensive, and embarrassed," and its cost is paid twice: in hours spent on buyers who could never afford the offer, and in buyers who never enquired and therefore never appear in any report.
• A quotation is built from ten blocks, and the three most often missing are the exclusions, the assumptions and the expiry date, which are also the three that produce the most expensive arguments later.
• Options are manufactured, not copied: the gap must be large enough to matter, small enough to climb and explainable by contents, and on the case the step to the top option is $600.00 a month, or 9.90 percent of a first year.
• The first price shown must be one the buyer can actually pay, and only amounts imposed by legislation may be added afterwards; the two Québec taxes add 14.975 percent to the case's pre-tax total of $54,144.00, for $62,252.06 payable.
• A deposit and a net period are pricing decisions: the case's deposit of $16,243.20 covers 64.46 percent of what the twelve kits cost to build and install, and thirty extra days on the balance costs $373.82.
• A quotation leaves four records — the document, the send log, the approval trail and the price evidence — and a business that keeps none of them cannot measure its own pricing, whatever tools it buys later.

Your Action Plan

1. Decide, offer by offer, whether to publish. The deliverable is a one-page list of everything the business sells, each line marked "publish," "publish a starting point" or "quote"; it takes about an hour, and it is done when every line carries a reason from Table 7.1 rather than a habit.
2. Audit every place your price is visible. The deliverable is a list of every page, document, listing and template showing a price, with the number each shows and the date it last changed; it takes about two hours, and it is done when every number is either current or corrected.
3. Rebuild your quotation template against the ten blocks. The deliverable is one template containing all ten blocks of Table 7.3, with the assumptions block and the validity date present by default; it takes half a day, and it is done when a colleague who was not involved can read a filled example and answer the five questions of the Method callout.
4. Write the exclusions and assumptions once, for reuse. The deliverable is a standard list of what the business does not include and what it assumes, with the consequence when an assumption fails; it takes about two hours, and it is done when the last three jobs that went over scope are each covered by a line.
5. Price your payment terms. The deliverable is a short table giving your cost of money, what each additional thirty days costs on a typical order, and the deposit you will ask; it takes about an hour, and it is done when the person who negotiates can state the cost of an extra thirty days from memory.
6. Set a validity period and enforce it for one quarter. The deliverable is an expiry date on every quotation and a sentence saying what follows it; it takes ten minutes to write and a quarter to prove, and it is done when at least one expired quotation has been reissued at current prices rather than quietly honored.
7. Start the quotation log. The deliverable is one sheet with a row per quotation sent, carrying the date, the buyer, the total, the option recommended, the option accepted, the discount and the outcome; it takes thirty minutes to create, and it is done when it has been filled for a full month, because Chapter 9 measures from it.

In La Machine

Every route named in this section was verified in the route inventory on September 19, 2026 and read in the product's source code at commit be8bf04 of September 7, 2026 (La Machine, 2026); nothing below is inferred from a page title.

• Business → Motivational Quotes (business/quotes) is a page of twenty-six hard-coded motivational quotations with a first-name personalization field and a set of categories, and it holds no commercial document of any kind.
• Business → Session Packages (business/packages, also reachable at productivity/packages) and business/subscription-packages hold the priced items a customer can be shown, each with a price and a cost, the subscription package carrying a billing period that may be monthly or annual and nothing else.
• Business → Store → Products (business/shop and business/shop-product-edit) holds a store product with a price, a currency, a status, a product cost, a shipping cost and shipping fees, and displays a profit panel computed from them on screen.
• client/packages and client/store display those prices to a customer, and client/packages leads to /checkout-demo, which shows a "DEMO MODE" banner and states that taxes will be calculated at the next step.
• What is not here. There is no quotation, no proposal, no estimate, no order form, no cart and no price list; the page at business/quotes holds motivational quotations. There is no invoice for the reader's own customers and no accounts-receivable module. There is no validity period, no expiry date, no option set, no deposit field, no payment term, no net period and no electronic signature, the only typed name in the platform being the one on the intake form. Taxes are never computed at payment: /checkout-demo announces taxes at the next step and never calculates them, offers no field for any code and debits nothing, and /client-subscribe is disabled; the only place a sales tax appears is the fictitious data generator behind Account → Profile → Revenues & Expenses (business/subscriptions). There is no price-page builder and no control over the prominence of one number against another, and the platform checks no price presentation against the Competition Act or the Consumer Protection Act, which Book B1, Chapter 6, states already and this book confirms.

Self-Check

1. What are the five conditions under which publishing a price is the right posture, and which one does the case's installed kit fail?
2. What does a hidden price communicate to a buyer, and what are the two costs of that communication?
3. Which blocks of a quotation are most often missing, and what does each of them prevent?
4. On the case's quotation for twelve buildings, what is the pre-tax total, what are the two taxes, and what is the total payable?
5. What does a deposit buy, and what does an additional thirty days of payment term cost on the balance of that quotation?

Answers. (1) A standard scope; a cost to serve that does not vary by more than the margin can absorb; no expected negotiation, or a list price with published reductions; the ability to honor the number everywhere it is visible; and the ability to keep the page current with an effective date. The kit fails the second, the installation cost varying materially from site to site (Section 7.1.3 and Table 7.1). (2) It says "expensive, and embarrassed," and suggests the price is a function of the buyer rather than of the offer; the two costs are the hours spent on enquiries from buyers who could never afford the offer, and the buyers who never enquired and therefore never appear in any report (Section 7.1.2 and the Key Concept callout). (3) The exclusions, which prevent unbilled work; the assumptions, which prevent a variance absorbed during delivery; and the validity date, which prevents a price being honored at a cost base that has moved (Sections 7.3.1, 7.3.3 and 7.6.3, and Table 7.3). (4) The pre-tax total is $54,144.00 for twelve kits at $4,512.00; the Goods and Services Tax is $2,707.20 and the Québec Sales Tax is $5,400.86; the total payable is $62,252.06 (Section 7.5.2 and Figure 7.1). (5) A deposit finances the outlay, qualifies the buyer, prices the risk of cancellation and signals symmetry; on the case it is $16,243.20, covering 64.46 percent of the $25,200.00 the twelve kits cost, and thirty extra days on the balance of $37,900.80 costs $373.82 at the model's 12 percent cost of money (Sections 7.6.1 and 7.6.2, and Table 7.6).

Chapter References

Competition Act, RSC 1985, c C-34, ss. 52, 74.01, 74.05 and 76 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Competition Bureau Canada. (2014, September 15; modified 2024, June 27). Price maintenance (section 76 of the Competition Act): Enforcement guidelines. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/en/how-we-foster-competition/education-and-outreach/price-maintenance-section-76-competition-act

Consumer Protection Act, CQLR c P-40.1, s. 224 (Québec). LégisQuébec, consolidation of April 7, 2026, as read for Book E1 on September 8, 2026; not read again, LégisQuébec returning an access error on September 19, 2026.

Consumer Protection Act, CQLR c P-40.1, ss. 54.1, 261 and 262 (Québec). LégisQuébec, consolidation of June 12, 2026, as read for Book H1 on September 13, 2026; not read again, LégisQuébec returning an access error on September 19, 2026.

Federal Trade Commission. (2025, May). The Rule on Unfair or Deceptive Fees: Frequently asked questions. Retrieved September 19, 2026, from https://www.ftc.gov/business-guidance/resources/rule-unfair-or-deceptive-fees-frequently-asked-questions

Guides Against Deceptive Pricing, 16 C.F.R. pt. 233 (2026) (United States). Canonical source https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-233; text read September 19, 2026, at the Legal Information Institute, Cornell Law School, which publishes no consolidation date.

Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Office de la protection du consommateur. (2025, February 10). Annonce d'un prix tout inclus. Gouvernement du Québec. Retrieved September 19, 2026, from https://www.opc.gouv.qc.ca/commercant/pratique-commerce/publicite-loi/prix-tout-inclus

Rule on Unfair or Deceptive Fees, 16 C.F.R. pt. 464 (2026) (United States) (source: 90 Fed. Reg. 2166, January 10, 2025; effective May 12, 2025). Canonical source https://www.ecfr.gov/current/title-16/chapter-I/subchapter-D/part-464; text read September 19, 2026, at the Legal Information Institute, Cornell Law School, which publishes no consolidation date.

Chapter 8 — Pricing Changes Over Time

At the start of its third year, Hydrosentinelle inc. was charging two different prices for the same monitoring service. New buildings went onto the ladder Chapter 5 built, which starts at $89.00 a month; the 206 buildings that had signed in the first two years were still paying $79.00 a month, the price the founders had set in year 1 and had never touched. Nobody had decided to run two price lists. The president, Laurence Bergeron, took the view that the first customers had built the company and should never see an increase. The cofounder responsible for sales and operations, Karim Haddad, took the view that they should all move to $89.00 on the first of the following month. What the company actually did is the subject of this chapter: sixty days of written notice, six months during which the old price continued, and then one step of $10.00 a month. Six subscriptions were lost beyond ordinary churn, two hundred were kept, and the move was worth $9,696.00 in year 3 and $19,392.00 in a full year. The arithmetic of Chapter 2 had said, before anyone wrote a letter, that the company could afford to lose 28 subscriptions and still be no worse off. It lost six. The gap between six and twenty-eight is the finding of this chapter, and it is not that the increase was risky. It is that the increase was late.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Recognize that a price left alone is a price falling, and name the four things that move underneath it while it stays still.
• Distinguish a legitimate reason to raise a price from a wish, including the case where capacity, not cost, is the binding constraint.
• Size an increase from evidence rather than from nerve, using the value ceiling, the alternatives and the price your own new customers already pay.
• Compute the tolerable loss before the increase, so that you know how many customers you can afford to lose and can measure the result against a number you set in advance.
• Choose a grandfathering period and defend it, knowing what the delay costs you and what a permanent legacy price costs you later.
• Write an indexation clause that can be applied without an argument, name the public index precisely, and decide whether to sell the buyer a cap.

Why It Matters

Most small businesses change their prices the way they change their insurance broker: once, under pressure, years after it was needed. The reason is not laziness but arithmetic that nobody has done. An increase feels like a bet on losing customers, and a bet with an unknown downside is one no sensible owner takes. This chapter makes the downside knowable. On the case, the break-even arithmetic of Chapter 2 says that a step from $79.00 to $89.00 on a subscription whose contribution becomes $74.00 can survive the loss of 13.51% of the base, which on 206 subscriptions is 28 of them. That is not a feeling; it is a threshold, computed before the decision and compared with the result afterwards. An owner who has that number behaves differently from an owner who does not.

For a solo business, the chapter matters because nobody else will ever raise the price. There is no committee, no annual review that happens by itself, and no colleague who notices that a rate set four years ago has quietly become a discount. For a small team, it matters because the increase is an operation with parts that must arrive in the right order: the notice, the price page, the quotation template, the invoice, and the answer the seller gives when the telephone rings. An increase that reaches the invoice before it reaches the website is not a communication error but a legal exposure, because supplying a product above the price advertised for it is reviewable conduct under the Competition Act. For an established business, it matters because legacy prices accumulate. Every grandfathered cohort is a permanent obligation, every uncapped multi-year contract is a commitment made at a price the business no longer charges, and the administrative cost of maintaining six price lists eventually exceeds the goodwill that created them.

There is a third reason, and it belongs to the whole of Collection A. Price is the strongest of the four profit levers, and Chapter 1 measures it on the case: one point of price is worth $11,303.88 of operating profit in year 3, which is 6.22% of the year's EBITDA, against $7,018.08 for a point of volume. A business that never moves its price is choosing, year after year, not to use the strongest lever it has. The full-year value of the single step taken in this chapter is $19,392.00 against a baseline EBITDA of $181,808.00 — more than a tenth of the year's operating profit, produced by editing one number and writing one letter.

In This Chapter

• 8.1 Why a Price That Never Moves Is Falling
• 8.2 The Legitimate Reasons to Raise a Price
• 8.3 Sizing the Increase
• 8.4 The Tolerable Loss
• 8.5 Grandfathering: Who Keeps the Old Price, and for How Long
• 8.6 Indexation and Multi-Year Contracts
• 8.7 Communicating an Increase
• 8.8 Lowering a Price, and When Not To

8.1 Why a Price That Never Moves Is Falling

8.1.1 A price is the only number in the business that waits to be edited

Everything a price is made of moves without being asked. Wages move, insurance moves, components move, rent moves, subcontractors move, and the value the offer creates for the buyer moves as the offer improves. The price alone sits in a document until a human being changes it. That asymmetry is the whole of this section: a business that takes no decision about its price has not held it steady, it has lowered it, by exactly the amount that everything underneath it has risen.

The movement is measured, and in Canada it is measured monthly. Statistics Canada's all-items Consumer Price Index rose 0.7% on an annual average basis in 2020, 3.4% in 2021 and 6.8% in 2022, which the agency described as a forty-year high and the largest increase since 1982, when the index rose 10.9%; within 2022 the twelve-month change accelerated to a peak of 8.1% in June before slowing in the second half (Statistics Canada, 2023). Four years later the pace is calmer: the all-items index rose 3.0% year over year in August 2026, matching the 3.0% recorded in July (Statistics Canada, 2026b). Two cautions belong beside those numbers before any owner uses them. The Consumer Price Index measures a fixed basket of goods and services as experienced by Canadian consumers; it is not a measure of any particular business's input costs, and a company whose costs are mostly electronics and skilled labor will find its own cost line moving at a different rate. And an index is a description of the past, not a license for a future price: the reason to raise a price is always the specific evidence set out in Section 8.2, and the index is only one of the places that evidence comes from.

DEFINITION Price increase, grandfathering, indexation and tolerable loss A price increase is a deliberate change to the price charged for an unchanged or improved offer, taking effect on a stated date and applying to a stated population of customers. It is distinct from a price change at renewal, which is the same act performed at a moment the contract already provides for, and from a re-pricing of a new offer, which changes what is sold as well as what is charged. Grandfathering is the practice of allowing existing customers to keep the price they were paying, either for a stated period or permanently, after a new price has been published for everyone else; the population that keeps the old price is called a legacy cohort, and the price it keeps is a legacy price. Indexation is a term written into a contract or a quotation by which the price moves automatically on a stated date, by reference either to a fixed percentage agreed in advance or to a named public index, with or without a cap, which is a maximum movement the seller accepts in exchange for the buyer's agreement to the mechanism. The tolerable loss is the number of customers, or the percentage of volume, that a business can lose after an increase and still earn the same total contribution as before; it is the break-even volume change of Chapter 2 applied to an increase rather than to a discount, and it is computed before the increase, not after. In plain terms: the increase is the decision, grandfathering decides who it reaches and when, indexation decides whether the next one has to be decided at all, and the tolerable loss decides whether the decision was sound.

8.1.2 The case is already running two price lists

Hydrosentinelle did not face a choice between raising its price and leaving it alone. It had already raised it. The ladder published at the start of year 3 charges $89.00, $129.00 and $179.00 a month for one building, and Chapter 5 shows that the mix of new buildings produces an average of $112.00 a month against the single old price of $79.00. Every building signed from that day forward pays the new prices. The 206 buildings signed before that day pay $79.00. The service they receive is the same service.

Two price lists for one service is a position, not a neutral state, and it has three costs that compound quietly. The first is the money, which Section 8.3 counts. The second is the administration: two prices mean two invoice templates, two renewal conversations and two answers to the same question. The third is the one that eventually forces the decision anyway, which is that the two populations meet. Property managers in one city belong to the same association, and Chapter 3 makes the point in its own context: in a market of a few hundred buyers, a price difference that cannot be explained is discovered. The explanation "you signed earlier" is a good one and it works for a while. The explanation "we never got around to it" is not an explanation.

8.1.3 What standing still costs, in the case's own numbers

The table that follows sets out what moves while a price stays where it is. Its first column names what moves, the second gives the measurement and its date or its model block, the third says where the movement lands in the accounts, and the fourth says what a business that never edits its price does about it. Read the fourth column as a single sentence repeated five times: the business absorbs the movement out of its own margin.

Table 8.1 — What moves while a price stays where it is

The money at stake is not small, and the case gives its size twice over. Chapter 1 measures the four profit levers on year 3 and finds that one percentage point on the price is worth $11,303.88 of operating profit, against $7,018.08 for a point of volume, $4,285.80 for a point of variable cost and $5,200.00 for a point of fixed cost. The single step this chapter takes is worth $19,392.00 in a full year, net of the customers it costs, on a baseline EBITDA of $181,808.00. No other decision available to this company in year 3 produces that much for two months of preparation and one letter.

WHAT CHANGED Consumer price inflation peaked in 2022, and the clauses written during it are still running Between 2020 and 2022 the rate at which Canadian consumer prices moved changed by an order of magnitude. Statistics Canada records annual average increases of 0.7% in 2020, 3.4% in 2021 and 6.8% in 2022, the last being a forty-year high and the largest since 1982, with a peak twelve-month change of 8.1% in June 2022 (Statistics Canada, 2023). By August 2026 the twelve-month change had returned to 3.0%, unchanged from July (Statistics Canada, 2026b). Three consequences follow for the subject of this chapter, and the first two are measured while the third is this book's reading of them. First, a multi-year price agreed before 2021 and never indexed was overtaken by a movement nobody had priced. Second, expectations have separated from costs: among Canadian businesses with one to nineteen employees surveyed in the second quarter of 2026, 24.3% expected to raise their selling prices over the next three months and 65.2% expected no change at all, while 64.0% expected rising input costs to be an obstacle (Statistics Canada, 2026a). A business that expects its costs to rise and its prices not to has described the erosion this chapter is about, in its own words, in a national survey. Third, and this is interpretation rather than measurement, the indexation clauses written in the middle of that episode were written under the impression of 2022 and are being applied in the conditions of 2026: sellers who accepted a cap without a floor discovered that the mechanism is asymmetric, and buyers who agreed to an uncapped index at the peak learned to refuse one. Section 8.6 draws the practical conclusions. Verify the current values of the index before writing any of them into a clause.

8.2 The Legitimate Reasons to Raise a Price

8.2.1 Six reasons, and the evidence each one requires

An increase is defensible when the business can state, in one sentence, what changed, and can produce the document behind the sentence. That is a higher standard than most increases meet and a lower one than most owners fear. There are six reasons that meet it. Costs have risen, and the business can show the cost line and the date it moved. The offer has changed, and the business can name what was added and when. The value the buyer receives has been measured and found larger than the price assumed, which is the work of Chapter 2. The price was wrong when it was set, and the evidence is that the market is paying more for the same thing today. Capacity is the binding constraint, and the business is selling more than it can deliver well. Or the business has repositioned, serving a different segment with a different offer, which is properly a new price rather than an increase.

The table that follows sets the six out. Its first column names the reason, the second states the evidence the business must hold before it writes to anybody, the third gives the sentence the buyer is actually told, and the fourth says where in this book the arithmetic sits. The third column is the one to draft first: a reason that cannot be said to the buyer in one true sentence is a reason the business has not finished thinking about.

Table 8.2 — Six legitimate reasons to raise a price, and what each one requires

Two of those six deserve a warning that the table cannot carry. The first is the cost reason, which is the most used and the weakest, because it invites the buyer to answer that their own costs have risen too and that the seller's inefficiency is not their problem. It works when it is specific, dated and modest, and it fails when it is a general appeal to inflation. The second is the improvement reason, which is strong when the improvement is visible and dishonest when it is not. A business that adds a feature nobody asked for in order to justify a step it had already decided on has not improved the offer; it has manufactured a pretext, and buyers who are paying attention can tell the difference.

8.2.2 Capacity: when delivery is the constraint, price is the honest rationing device

Book E1 (The Sales Process: From Prospect to Signed Contract), Chapter 10, sends this book a case that most pricing writing ignores. When capacity is the binding constraint — when the business is already selling more than it can deliver well — the remedies lie on the delivery side, in Book J1 (Operations Management Fundamentals), or in the offer, by raising prices, and the correct sales response is to raise the qualification bar rather than to sell harder. This chapter supplies the pricing half of that answer, and it begins with an observation Chapter 1 also makes: a business that is fully booked and still quoting at the same price is not being competitive, it is rationing by queue.

Rationing by queue is rationing by patience. The customer who waits is the one with the least urgent problem, not the one who values the work most, and the business ends up serving the least valuable half of its demand at the same price as the most valuable half. Rationing by price does the opposite, and it pays for the capacity that would relieve the constraint. The evidence a business should hold before using this reason is operational rather than financial: utilization, lead time from order to installation, the number of opportunities declined or deferred, and the overtime being worked to keep promises. When lead time is growing and work is being refused, the price is too low, whatever the cost line says.

Three cautions keep this reason honest. A capacity increase is legitimate while the constraint lasts, which means the business must decide in advance what it will do when the constraint lifts: either invest in capacity and hold the new price because the offer now reaches more buyers, or accept a smaller and better-paid book of business. Deciding nothing produces the worst outcome, which is a price raised in a busy quarter and quietly discounted in a slow one, teaching every buyer that the price depends on the seller's calendar. The second caution is that a capacity increase should never be dressed as a value increase; buyers who are told the offer improved and find that it did not remember. The third is that the increase should be applied to new work first. Raising the price on customers already waiting for delivery converts a capacity problem into a trust problem, and the queue is full of people who will talk to each other.

8.2.3 Three reasons that are not reasons

Three arguments turn up in every increase discussion and none of them survives contact with a buyer. The first is that the business needs the money. A budget gap is real and it is the seller's problem, not the buyer's, and an increase justified that way invites the buyer to negotiate over the seller's finances rather than over the value received. The second is that everyone else has raised. Following a competitor's published move is lawful and ordinary; agreeing with a competitor to move together is a criminal offense under section 45 of the Competition Act, as Chapter 2 sets out, and the word "price" in that section expressly includes rebates, allowances and concessions. The safe conduct is to observe published prices and decide alone, and to leave any meeting where the subject is raised. The third is that the customer can afford it. Ability to pay is not a reason and is not defensible if it is ever said aloud; what is defensible is that the value delivered to that customer is larger, which is a different statement and requires the measurement of Chapter 2 to support it.

WATCH OUT Raising the price without changing anything the customer can see An increase applied to an offer the buyer perceives as identical is the most common increase and the most expensive. It is not unlawful and it is often justified, but it places the entire weight of the decision on the notice letter, and the letter has to carry an argument the customer cannot verify. Three failures follow. Buyers with alternatives use the moment to re-open the whole decision, so a step of a few dollars triggers a competitive review that would otherwise not have happened for years. Buyers without alternatives pay and remember, and the memory is spent at the next renewal, on the next quotation, and in the conversation they have with the next buyer who asks them for a reference. And the seller, sensing all this, begins granting exceptions on the telephone, which converts a clean increase into a patchwork of individual prices that costs more to administer than the increase earned. The remedy is not to invent an improvement. It is to time the increase so that it lands beside something real: a feature that shipped, a service level that was raised, a response time that was shortened, a reporting pack that was added. On the case the timing was deliberate, because the $89.00 entry price is the price of a ladder whose upper tiers carry genuinely more service, priced at $31.00 and $38.00 a month to serve against $15.00 for the entry tier, so an existing customer being moved to $89.00 is being moved onto a published structure rather than being charged more for nothing. Where no such anchor exists, say plainly that costs have risen, give the date of the last increase, and keep the step small enough that the sentence is credible.

8.3 Sizing the Increase

8.3.1 Four inputs decide the size, and only one of them is the cost

The size of an increase is bounded above by the value the buyer receives and below by the cost of serving them, and the four inputs that locate it between those bounds are, in descending order of usefulness: the price the business already charges new customers, the value measured for the buyer, the prices of the alternatives the buyer can actually buy, and the movement in the business's own costs since the last change. Most owners use them in the reverse order, starting with the cost movement because it is the only one they have on file, which is why most increases are both too small and hard to defend.

On the case the first input settles the question. New buildings have been paying $89.00 a month since the start of year 3, and they are buying the same service as the 206 buildings paying $79.00. The evidence that $89.00 is acceptable is the company's own order book, which is stronger evidence than any estimate, any competitor comparison and any cost calculation, and it has a further advantage: it produces a target price that the business can explain in a single sentence without arguing about inflation. The second input confirms that the ceiling is far above the step. Chapter 2 measures the economic value the service creates for a forty-eight-unit rental building at $5,668.00 a year against the $1,908.00 the case charges, so the company captures 33.66% of the value it creates. Moving the entry subscription by $10.00 a month changes that share by a little; it does not approach the ceiling. The third input says the same thing from the outside: the annual costs of ownership of the three competitors Chapter 2 examines are $1,368.00, $2,308.00 and $2,788.00, so the case sits in the middle of its market before the step and remains there after it.

8.3.2 One step or a staircase

Having chosen a destination, a business must choose how to arrive. A single step is cheaper to administer, easier to explain and finished sooner; a staircase of smaller steps is easier for a buyer to absorb and keeps the seller in the habit. The choice turns on three questions. How long has it been since the last change? A business that has not moved in three years has a gap to close and should close it in one step, then adopt a cadence, because a business that closes a three-year gap in three annual installments spends three years explaining increases. What does the buyer's own budget cycle look like? A buyer who sets an annual budget in November will absorb a step announced in September and resent the same step announced in January. And what does the step do to the tolerable loss? A larger step buys a larger cushion, which is the counter-intuitive result of Section 8.4 and the reason that timid increases are often the riskiest ones: they cost the same in goodwill and produce less to show for it.

The rule this book recommends for a small business is a catch-up followed by a cadence. Close the gap once, with proper notice and a grandfathering period, then review prices on a fixed date every year, whether or not anything has changed, so that the review is a routine rather than an event. Chapter 10 places that review in the calendar with the rest of the pricing cadence. The annual review does not oblige the business to raise anything; it obliges the business to decide, which is the thing that was not happening.

8.3.3 The step the case actually took

The step is $10.00 a month, from $79.00 to $89.00, which is an increase of 12.66%. It is large by the standard of a routine annual adjustment and small by the standard of the gap it closes, since it does nothing more than bring existing customers to the price that new customers already pay. It was announced 60 days in advance and took effect after 6 months of grandfathering, which are the two parameters Section 8.5 and Section 8.7 examine. The arithmetic below is the whole of the financial case, and it is deliberately simple enough that a property manager could check it.

DO THE MATH Raising seventy-nine dollars to eighty-nine on 206 subscriptions Fictional example; the prices, the churn assumption and the timing are illustrative assumptions of the model. Start with the population: 206 subscriptions at $79.00 a month at the beginning of year 3, each costing $15.00 a month to serve, so each contributing $64.00. The new price is $89.00, a step of $10.00 and an increase of 12.66%, and the new contribution per subscription is $74.00. Step one, the customers lost. The model assumes that an increase of this size costs 3% of the base beyond the churn that would have happened anyway, which on 206 subscriptions is 6 subscriptions; 200 are kept. Step two, what the increase earns in year 3. Six months of grandfathering mean the new price is billed for 6 of the 12 months of year 3, so the gain is 200 subscriptions times $10.00 times 6 months, which is $12,000.00. Step three, what the lost customers cost. The 6 subscriptions that leave were each contributing $64.00 a month, and over the same 6 months that is $2,304.00 of contribution forgone. Step four, the net. $12,000.00 less $2,304.00 is $9,696.00 in year 3. Step five, a full year. With the grandfathering period behind it, the same arithmetic over twelve months gives 200 times $10.00 times 12, which is $24,000.00, less 6 times $64.00 times 12, which is $4,608.00, for $19,392.00. Step six, the threshold. The break-even arithmetic of Chapter 2 says the company can lose 13.51% of the base before the increase stops paying, which on 206 subscriptions is 28 subscriptions. The assumption is six. The threshold is twenty-eight.

The table below restates that arithmetic line by line, so that a reader can lift it into a spreadsheet. Its first column names the line, the second gives the figure, and the third says what the figure is and where it comes from. The last two lines are the ones to compare with each other, because the whole judgment of the increase lies in the distance between them.

Table 8.3 — The increase on the case, line by line

8.4 The Tolerable Loss

8.4.1 The question that should be asked before the letter is written

The question every owner asks before an increase is "how many customers will I lose?", and it cannot be answered, because it is a forecast about human behavior in a market of a few hundred people. The question that can be answered is the other one: "how many customers could I lose and still be no worse off?" That number is arithmetic, it takes two minutes, and it converts a bet into a decision with a stated threshold.

The rule in words is short. For a price increase, the tolerable percentage loss of volume equals the price change divided by the contribution per unit after the change. On the case, the step is $10.00 and the contribution after the step is $74.00, because the new price of $89.00 carries a cost to serve of $15.00. Ten divided by seventy-four is 13.51%, and 13.51% of 206 subscriptions is 28. Everything else about the increase — the notice, the letter, the telephone calls, the grandfathering — is a question of how far below twenty-eight the business can hold the actual loss.

The same rule read backwards explains why price decreases are so much harder than increases, and Chapter 2 establishes the point in full. For a decrease, the required percentage gain in volume equals the price change divided by the contribution per unit after the change, and because a decrease makes the denominator smaller while the numerator stays the same, the required gain is always larger than the tolerable loss for an equivalent increase. An increase gives the business a cushion. A decrease demands a result.

8.4.2 Six against twenty-eight

The model assumes that a step of this size costs 3% of the base beyond ordinary churn, which is 6 subscriptions. That assumption is not a measurement, and a reader should treat it as this book treats every behavioral assumption: as an input that travels with its result. What makes the case persuasive is not the assumption itself but the distance between it and the threshold. The extra churn would have to be nearly five times larger than assumed before the increase stopped paying for itself, and it would have to be more than four times larger before the company was worse off than doing nothing.

Three features of this particular offer support the low assumption, and each one is a question a reader should ask about their own. The first is switching cost: the sensors are installed in the building, and replacing them is a capital project rather than a decision, which means the buyer's alternative to paying ten dollars more a month is an installation. The second is the size of the step relative to the buyer's own economics: ten dollars a month for a building whose exposure Chapter 2 measures in thousands is not a budget event. The third is the alternative that exists: Chapter 2 records a competitor whose monitoring is $49.00 a month, so the increase makes a cheaper option relatively more attractive, and the buildings most likely to leave are the ones that were closest to choosing it in the first place. A business raising a price should be able to name the customers most likely to leave before the letter goes out, because those are the calls that will come.

There is also a piece of outside evidence that cuts against the intuition behind the fear. Retention data gathered in 2025 across roughly 3,500 subscription software companies found that retention separates sharply by price level rather than falling with it: products priced above $250 a month showed a median gross retention of 70% and net revenue retention of 85%, while products below $50 a month showed 23% and 32% (ChartMogul, 2025). The categorization in that study was assisted by artificial intelligence reading company websites, and the sample is a provider's own customer base rather than a probability sample, so the figures describe that population and no other. Taken for what they are, they say that a low price does not buy loyalty, and that the customers most likely to leave are not the ones paying the most.

There is also an honest reading in the other direction. Where switching costs are low, where the offer is undifferentiated, where the buyer buys on price and where a substitute is one click away, an increase of 12.66% may cost far more than 3% of the base, and the tolerable loss may be reached. The correct response is not to abandon the increase but to change the order of operations: raise the price on new customers first, watch the win rate for two quarters, and only then move the existing base. That is precisely the sequence the case followed without planning it, which is why the evidence for the step was already in the order book by the time the letter was written.

8.4.3 Four moves, and the volume each one can survive

The table below applies the break-even arithmetic to four moves the case could make, two upward and two downward. Its first column names the move, the second gives the change in price, the third gives the break-even change in volume, and the fourth says how to read the result. A negative figure in the third column is a loss the business can absorb; a positive figure is a gain the business must achieve. The contrast between the two halves of the table is the strongest argument in this chapter.

Table 8.4 — Four price moves on the case, and the volume change each one can survive

Two readings follow, and the second is the one most owners have never seen. The first is that increases are structurally safer than decreases: the kit can lose 8.16% of its volume on a 5% increase, while a 10% cut on the same kit demands 21.62% more volume, and a business that cannot name where that volume would come from should not make the cut. The second is that the cushion grows with the size of the step. The subscription can lose 13.51% of its base on a 12.66% increase, which is more tolerance than the kit gets for a 5% increase, because the subscription's contribution margin is far higher. High-margin offers can absorb more customer loss per point of price than low-margin offers, which is exactly the opposite of what fear suggests, and it means that the offers a business is most afraid to re-price are usually the ones it is safest to re-price.

Figure 8.1 puts the whole decision on one page: the 206 subscriptions before the increase, the 6 lost beyond ordinary churn, the 200 kept, the money on the right — a gain of $12,000.00 in year 3 against $2,304.00 of contribution forgone, a net of $9,696.00 and $19,392.00 in a full year — and, drawn as a dashed line far beyond the six, the tolerable loss of 28 subscriptions.

Figure 8.1 — The increase from seventy-nine to eighty-nine dollars, against the loss it could have afforded. The left panel shows the population: 206 subscriptions before the increase, 6 lost beyond ordinary churn under the model's assumption of 3%, and 200 kept, with the tolerable loss of 28 subscriptions drawn as a dashed line far to the right of the six actually assumed. The right panel shows the money: a gain of $12,000.00 over the six months of year 3 that follow the six-month grandfathering period, $2,304.00 of contribution forgone on the six subscriptions lost, a net of $9,696.00 in year 3, and $19,392.00 in a full year at the new price. The step is $10.00 a month, an increase of 12.66%, announced 60 days ahead. Hydrosentinelle inc. is fictional, and the prices, the churn assumption and the timing are illustrative assumptions of the model.

8.5 Grandfathering: Who Keeps the Old Price, and for How Long

8.5.1 What a grandfathering period buys, and what it costs

Grandfathering is the seller's way of buying time with money. It buys three things. It buys goodwill, because a customer told that the new price applies in six months is being given something the seller did not have to give. It buys sequencing, because the cancellations that do come arrive spread over months rather than in a single week, which matters to cash and to the nerves of the person answering the telephone. And it buys a defensible story, because "we published a new price in January and moved existing customers in July, with sixty days' notice" is a sentence that survives being repeated by a customer to another customer.

What it costs is visible in the case and is exactly half the first year. The step earns $12,000.00 in year 3 instead of the $24,000.00 that a full year at the new price would produce, because the new price is billed for 6 of the 12 months. That is the price of the six months of goodwill, and it is worth stating in those terms when the decision is taken, because a business that has not priced its own generosity tends to extend it. The second cost is administrative: for the length of the period the business is running two prices deliberately rather than by neglect, and every invoice, every renewal and every quotation has to know which population the customer belongs to.

The third cost arrives later and is the one that ruins companies slowly. A grandfathering period that is not dated becomes a legacy price, and a legacy price never ends. It compounds with the next increase, because the legacy cohort is now two steps behind, and by the third increase the business is maintaining three price lists and cannot answer the question "what do we charge for this?" without asking when the customer signed. The discipline is simple and it is entirely in the wording of the letter: a grandfathering period has an end date in the first sentence that announces it.

8.5.2 Four ways to treat existing customers

The table that follows sets out the four treatments available. Its first column names the option, the second states what the existing customer keeps, the third states what the option costs the seller, and the fourth says when to choose it. The second and third options are the ones a small business will use; the first is for a business with no contracts and no relationships to protect, and the fourth should be chosen deliberately or not at all.

Table 8.5 — Four ways to treat existing customers when a price rises

The fourth option deserves its own sentence because it is chosen by accident more often than on purpose. A permanent legacy price is defensible in exactly one situation: a small, named, closed list of customers to whom the business owes something specific, counted and recorded, with the cost of the exception written down. It is indefensible as a general policy, and it is invisible as a general practice, which is how it becomes one.

8.5.3 Prepaid years, contracts, and the date the price actually changes

Three populations do not follow the general rule, and a business that has not listed them before writing the letter will discover them afterwards, one telephone call at a time.

The first is customers who have prepaid. The case sells a twelve-month prepayment at $853.20, which is twelve months of $79.00 at a 10% discount, and Chapter 4 treats that discount as a rate of interest the business pays for the cash. A customer who has prepaid has bought grandfathering, and the business must honor it to the end of the prepaid period; the increase applies at the next renewal of the prepayment, and the renewal notice is where it is announced. It is worth noticing that this is the single cleanest moment in the whole subscription to change a price, because the buyer is already making a decision on a date both parties know.

The second is customers under a contract with a fixed term. The price changes at the end of the term, not during it, unless the contract itself provides otherwise — which is what an indexation clause does, and why Section 8.6 exists. Book H1 (Contracts and Commercial Agreements) drafts the clause; this chapter decides the number that goes in it.

The third population is consumers, and here this book must be exact about what it knows. Hydrosentinelle sells to building owners and property managers, so most of its agreements are commercial, and the general common-law and civil-law position is that a price may be changed for the future on reasonable notice unless the contract says otherwise. A business selling to consumers in Québec is in a different regime. What this book can state from a reading it has verified is that a merchant may not charge a consumer a price higher than the advertised one (Consumer Protection Act, s. 224, as read for Book E1 on September 8, 2026, at the consolidation of April 7, 2026), which means that an increase must reach the advertising before it reaches the invoice. What this book cannot state is the content of the provisions of that Act governing the unilateral amendment of a contract of successive performance: LégisQuébec and the public case-law database returned an access error on September 19, 2026, and the text was not read, so nothing is quoted or paraphrased from it here. A reader who sells subscriptions to consumers in Québec should treat that as an open question to put to counsel, and should read Book H1 before drafting the clause rather than after sending the letter.

CASE IN POINT The sixty-day notice and the six-month grandfather Fictional example; names and figures are invented. The decision took one meeting in January of year 3 and turned on two numbers the founders had computed beforehand. The first was the threshold: 28 subscriptions, the loss the step could absorb. The second was the cost of waiting: $12,000.00 of the year's gain, which is what six months of grandfathering costs when the step is $10.00 a month on 200 kept subscriptions. The president, Laurence Bergeron, wanted twelve months of grandfathering; the cofounder responsible for sales and operations, Karim Haddad, wanted none. They settled on six months and 60 days of notice, for three stated reasons recorded in the pricing file. The first is that sixty days spans a full billing cycle with time left over, so no customer discovers the new price on an invoice. The second is that six months covers the annual budget cycle of a property manager, so no customer has to find the money in a budget that was already set. The third is that six months is short enough to be stated as a date rather than as a promise. The sequence was written down before anything was sent: the twelve largest accounts were telephoned by Karim Haddad in the week before the letters, every remaining customer received the same letter on the same day over the signature of the president, the published price page and the quotation template were changed on that day, and the invoice template was changed on the effective date and not before. Six subscriptions were lost. The founders recorded the result beside the threshold in the pricing file, with one line: next increase, review after twelve months, and do not wait three years again.

TOOLS OF THE TRADE What a billing capability must do to run an increase An increase is an administrative operation before it is a commercial one, and most of the pain small businesses report comes from tools that cannot express it. Seven functions are the requirements to bring to any comparison, and Chapter 12 compares the categories of tool that provide them, without prices. Effective dates: a price must be storable with the date on which it starts, so that the new price and the old price coexist and the system knows which is billed this month. Price history: every price a product has carried, with the dates and the person who changed it, because the ordinary-price questions of Chapter 6 and the disputes of Section 8.7 are answered from that record and from nowhere else. Cohorts: customers must be groupable by the price they are on, so that a grandfathered population can be listed, counted, invoiced correctly and moved on one date. Proration: a change taking effect mid-cycle must produce a correct partial charge without a manual credit note. Scheduled changes: the increase should be entered once, in advance, and take effect by itself, because an increase that depends on somebody remembering will reach the invoice before it reaches the price page. Notice and consent records: the date each customer was notified, by what means, and any acknowledgment, held with the account. Reporting on the result: the count of customers before, the count kept, the count lost, and the revenue effect, so that Section 8.4's threshold can be compared with what happened. A business whose tools provide none of these can still run an increase, from a spreadsheet and a mail merge, but it must keep the spreadsheet, because it is the only record it will have.

8.6 Indexation and Multi-Year Contracts

8.6.1 An unindexed multi-year price is a discount nobody negotiated

Book E1, Chapter 7, states the point in one line, in the row of its proposal table that deals with price, indexation and taxes, and this book's job is to put a number on it. The line is: an unindexed multi-year price is a discount nobody negotiated. The reason it is a discount is that a price fixed for three years is a price the seller has agreed not to change while everything underneath it changes, and the reason nobody negotiated it is that the buyer did not ask for it, did not pay for it, and in most cases did not notice it was being given.

KEY CONCEPT An unindexed multi-year price is a discount nobody negotiated Put the two numbers of this chapter beside each other and the sentence stops being rhetoric. On the case, one step of 12.66% on the subscription base was worth $19,392.00 in a full year, net of the six customers it cost. A business that signs a three-year agreement with no indexation has committed, in advance, not to take that step for three years. The first year of the freeze is worth nothing, because the step would not have happened yet. The second year is worth one step forgone. The third year is worth a second step forgone on top of the first, because the increase the business did not take in year two is still not taken in year three. In the arithmetic of this case, a three-year freeze is three such concessions, given without being asked for and without anything received in exchange — while the buyer, who would gladly have paid for the certainty if it had been priced, receives it for nothing. The same point in the case's largest quotation: Chapter 7 prices the recommended option for the Groupe immobilier Vallières at $1,548.00 a month for twelve buildings, which over a thirty-six-month term is $55,728.00 of monitoring revenue at a price fixed on the day of signature, while every cost inside it moves. The remedy is not to refuse long terms, which are valuable, but to charge for what a long term actually gives the buyer. A fixed multi-year price is a product: it is certainty, the buyer wants it, and it should have a price.

8.6.2 The three forms, and what a cap really is

A multi-year price can move in three ways, and a contract that provides for none of them has chosen the fourth, which is not to move at all. The first is a fixed percentage applying on each anniversary, agreed when the contract is signed. The second is a named public index, with the price moving by the change in that index over a stated period. The third is an index with a cap, where the price moves by the index up to a stated maximum, and it is the form most buyers accept because it converts an unknown into a bounded unknown.

The table below sets the three forms out. Its first column names the form, the second says how the clause is written, the third says what the form protects, the fourth states the buyer's usual objection, and the fifth says when to prefer it. A fourth line is included for completeness, because a contract that says nothing has made a choice and should be shown as having made it.

Table 8.6 — Three forms of indexation, and the one that is chosen by saying nothing

A cap is not a concession, it is a product, and it should be understood as one on the seller's side of the table. The seller is selling the buyer protection against an index that might move more than the cap, and the seller is bearing that risk. Two consequences follow. A cap should be paid for, most simply by setting the base price a little higher or by lengthening the term, and a seller who gives a cap for nothing has repeated the mistake of the unindexed price on a smaller scale. And a cap without a floor is asymmetric: the price cannot rise above the cap but it can fall all the way with the index. A collar — a cap and a floor together, most often a floor of zero, so that the price never decreases — is the symmetric version, and it is the version a seller should propose.

8.6.3 Naming the index so the clause can be applied without an argument

A clause that says "adjusted for inflation" is a future dispute. Inflation is measured by several series, for several geographies, on several bases, with revisions, and two reasonable people reading "inflation" will produce two different numbers. Six things have to be named in the clause for it to be applied mechanically: the publisher, the series, the geography, the basis, the reference period and the fallback. Name them, and an administrative assistant can compute the new price in five minutes; omit any one of them, and the annual adjustment becomes an annual negotiation, which is exactly what the clause was supposed to prevent.

QUÉBEC & CANADA NOTE Naming a Canadian public index in an indexation clause The usual index in a Canadian contract is the Consumer Price Index published by Statistics Canada, which measures price change by comparing, over time, the cost of a fixed basket of goods and services as experienced by Canadian consumers. Six elements make the clause mechanical. The publisher and the series: Statistics Canada, Consumer Price Index, all-items, published monthly in the data table numbered 18-10-0004-01, Consumer Price Index, monthly, not seasonally adjusted (Statistics Canada, 2026c). The geography: Canada, or a named province, or a named census metropolitan area — the same table publishes Canada, the provinces and territories, and metropolitan areas including Montréal and Québec City, and a Québec seller supplying Québec buildings may reasonably prefer the provincial series to the national one, provided it says which. The basis: not seasonally adjusted, which is the series contracts normally use, and the twelve-month change rather than the monthly change. The reference period: a named month, chosen far enough before the anniversary to have been published, because the index for a month appears in the middle of the following month — the August 2026 index was released on September 14, 2026 (Statistics Canada, 2026b). The rounding and the floor: how many decimal places, and whether a negative change reduces the price or leaves it unchanged. The fallback: what happens if the series is discontinued, renamed or rebased, which is a real event over a five-year term. Two further points belong to a Québec seller. Where the buyer is a consumer rather than a business, the increase must reach the advertising before it reaches the invoice, because a merchant may not charge a consumer more than the advertised price (Consumer Protection Act, s. 224, as read for Book E1 on September 8, 2026, at the consolidation of April 7, 2026). And the drafting of the clause itself belongs to Book H1 (Contracts and Commercial Agreements), not to this book, which decides only the number and the mechanism. Verify the current text and the current published values before relying on them.

8.6.4 The renewal, and the price a renewal offer carries

Book B1 (Marketing Fundamentals for Business Owners), Chapter 2, counts the renewal among the offers that sit at each stage of a customer journey, alongside the trial, the demonstration, the quotation, the package and the promotion. Chapter 4 of this book priced the first three and left the renewal here, because a renewal is an increase with a date on it. It is the one moment when a price can move without a special announcement, since the buyer is already deciding, the conversation is already scheduled, and neither party needs to explain why the subject has come up.

A renewal offer can carry three prices, and the choice is not neutral. It can carry the same price, which reads as a courtesy and is in fact a decrease in real terms, repeated every year until someone notices. It can carry the current list price, which is an increase for every customer who bought below list and is the default this book recommends, because it keeps one price list rather than a museum of old ones. Or it can carry a negotiated renewal price, lower than list, which is legitimate only under Chapter 6's rule of exchange: the reduction is bought with something — a longer term, an annual prepayment, a larger scope, a reference, a committed volume — and never given to end a silence.

Three practices make renewal pricing work. The first is that the renewal is announced before it happens, with enough notice for the buyer to decide, never applied silently on an automatic renewal date; an automatic renewal at a higher price that the customer discovers on an invoice is the fastest way to turn a small increase into a cancellation and a complaint. The second is that the renewal offer is prepared from the current price book and not from the last invoice, which requires the price book to exist. The third is that the renewal is used to improve the shape of the revenue as well as its level. Aggregated billing data from more than 2,500 subscription software companies over the full year 2024 found that annual plans retained better than monthly ones — at an average revenue per account between $250 and $500, net revenue retention was 88% for annual plans against 76% for monthly ones, and at low revenue per account, customer retention was 62% for annual against 41% for monthly, a gap of 21 points (Faustino, 2025). The same body of data found that the second month of a subscription is the peak moment of conversion from monthly to annual, customers being about three times more likely to switch then than in the ninth month. Both figures come from a provider's own customer base rather than from a probability sample, so they describe that population and not the whole market, and a small business should read them as a direction rather than as a forecast. The direction is clear enough to act on: a renewal conversation that moves a monthly customer to an annual commitment is usually worth more than the discount it costs, and it is worth more than a small price increase applied to a monthly customer who then leaves.

8.6.5 The clause belongs in the quotation, not only in the contract

Chapter 7 makes a point this chapter must not undo: the indexation choice appears in the quotation, because the quotation is the document the buyer compares and approves, and a mechanism that first appears in the contract sent for signature reads as a change of terms. The division of labor across the library is now complete for this subject. Chapter 5 sets the structure the price sits in. Chapter 7 presents the price and states the mechanism in the document the buyer reads. This chapter chooses the mechanism and sizes the movement. Book H1 (Contracts and Commercial Agreements) drafts the clause and makes it enforceable. Book G2 (Cash Flow Management and Forecasting) manages the cash consequences of long terms and prepayments, and this book does not repeat that work.

8.7 Communicating an Increase

8.7.1 The notice: how long, from whom, in what form

A price increase is announced in writing, individually, by the owner, with enough notice that no customer learns of it from an invoice. Sixty days is the figure the case used and a reasonable default for a monthly subscription, for three reasons that generalize. It spans a full billing cycle with time left over, so the customer sees the announcement, then one more invoice at the old price, then the change. It gives the seller time to speak to the largest accounts before the letter arrives, which is the single practice that most reduces the damage. And it is long enough to be courteous without being so long that the customer forgets, which is the failure mode of a six-month notice given without a reminder.

Two elements of form matter more than their cost suggests. The notice is individual, addressed to the customer, rather than a banner on a website or a line in a newsletter, because a customer who can say "nobody told me" is a customer who is right. And it is signed by the person who owns the price, which in a small business is the owner. Nagle et al. (2023) organize the subject so that price and value communication is a chapter of its own, sitting between the structure of a price and its level, which is a useful reminder of the order of operations: the level is decided first, in private, and the communication is designed afterwards, as a separate piece of work, rather than being improvised in the sentence that announces the number.

8.7.2 What the letter says, and what it must never say

A price increase letter has five jobs and should do nothing else. It states the new price and the exact date it takes effect. It states the old price, so that the customer does not have to look it up. It gives one reason, which is true and specific. It says what is not changing, which is usually most of the offer and is the part the customer cares about. And it says what the customer can do: prepay, commit to a longer term, move to a smaller tier, or call a named person at a named number.

Four things must not appear. An apology must not appear, because an apology invites a negotiation and signals that the sender does not believe in the number. A general appeal to inflation must not appear in place of a specific reason, because it is the sentence every buyer has read a hundred times. A promise never to raise prices again must not appear, because it will be broken. And an exception must not appear in the letter itself; exceptions are granted, if at all, one at a time, on the telephone, within the authority Section 8.7.3 sets out, and never offered to a population that has not asked.

The order of operations matters as much as the wording, and it is where increases most often go wrong. The correct sequence is: the largest accounts are telephoned first; every remaining customer receives the same letter on the same day; the published price page, the price list and the quotation template change on that day; and the invoice changes on the effective date and not a day earlier. Reversing any two of those steps creates a window in which the business is charging one price and advertising another, which is not merely embarrassing.

THE RULE Competition Act, s. 74.05: a business may not supply above the price it advertised Under the Competition Act, read on the Justice Laws Website, current to July 21, 2026, and last amended March 26, 2026, it is reviewable conduct for a person who advertises a product for sale or rent in a market to supply that product, during the period and in the market to which the advertisement relates, at a price higher than the price advertised (Competition Act, s. 74.05(1)). The exceptions in the provision are narrow: a catalog that prominently states that its prices may be in error, where the advertiser proves the error; an advertisement immediately followed by a correction; securities; and supply by a person who is not in the business of dealing in the product (Competition Act, s. 74.05(2)). The market covered is the one the advertisement could reasonably be expected to reach, unless the advertisement expressly restricts it to a stated area, store, department or catalog (Competition Act, s. 74.05(3)). Scope: federal law applying across Canada, in business-to-business sales as well as consumer sales, on the civil track, so no intent need be proved and the administrative monetary penalties of section 74.1 apply, reaching on a first order the greater of $10,000,000 and three times the benefit derived for a corporation, with a due diligence defense available to a person who establishes that reasonable care was taken to prevent the conduct (Competition Act, s. 74.1). This is the provision a price increase breaches by accident, and it breaches it in one specific way: the invoice is changed before the website, the price list, the catalog or the quotation template. The operational rule is therefore that the advertising changes first and the invoice changes last, and that every place a price is published is on a written list that is worked through on the same day. Verify the current text before relying on it.

8.7.3 What to do when the telephone rings

Some customers will call, and the number who call is smaller than the number who were going to leave anyway. The conversation has a shape. Acknowledge the increase without apologizing for it. Restate what the service does for that specific customer, in that customer's own numbers where they exist, which is the value estimate of Chapter 2 applied to one building. Then offer the exchange rather than the exception: a longer term at the old price, an annual prepayment, a move to a tier that genuinely contains less, or a phased step. Every one of those gives the business something in return, and every one of them is a decision the business can repeat consistently. An exception gives the business nothing and cannot be repeated consistently, because the next customer will hear about it.

Authority should narrow during an increase, not widen. Chapter 6 sets the case's standing limits at 6% for a representative and 12% for the head of sales, with a floor of $3,500.00 on the kit, which is 27.08% below list. During the sixty days of an increase, the sensible policy is that no representative may grant an exception at all and that every exception granted by the owner is recorded with the customer's name, the reason, the value of the exchange received and an end date. The reason is arithmetic rather than authoritarianism: the increase is worth $10.00 a month per subscription, so a handful of exceptions granted in the first week can cost more than the increase earns, and nobody will notice until the year is over. Book E1, Chapters 7 and 8, conducts the conversation itself, including the defense of a price against an objection; this chapter supplies only the boundaries inside which that conversation happens.

WHO DOES THIS Who writes the increase letter, and who answers the telephone In a solo business, one person does all of it, and the risk is not incapacity but sequence: the owner should write the letter on one day, with the threshold of Section 8.4 and the value estimate of Chapter 2 in front of them, and send it on another, because a letter written and sent in the same hour tends to contain an apology. In a small team, the division is clean and it matters. The owner decides the price, signs the letter and grants any exception. The person responsible for sales telephones the largest accounts in the week before the letters go out, in descending order of revenue, and hears the objections first. The person responsible for administration changes the price page, the price list, the quotation template and, on the effective date, the invoice template, and keeps the list of where prices are published. At Hydrosentinelle the president, Laurence Bergeron, signed the letter and took every call that asked for an exception, and the cofounder responsible for sales and operations, Karim Haddad, made the twelve calls that preceded it. In an established business the separation Chapter 11 describes applies with more force during an increase than at any other time: the person whose pay depends on closing this quarter's deals should not be the person who decides which customers are exempt from the new price. Two roles are needed whatever the size of the business, and both are often missing. Someone must own the list of every place a price is published, because section 74.05 is breached by the page nobody remembered. And someone must count the result against the threshold ninety days later, because an increase that is never measured teaches the business nothing about the next one.

METHOD Running a price increase in ten steps When to use it: whenever a price is to be raised for existing customers, and as the standing procedure attached to the annual price review of Chapter 10. Steps: (1) Establish the destination price from evidence, preferring, in order, the price new customers already pay, the measured value of Chapter 2, the alternatives available to the buyer, and the movement in your own costs. (2) Compute the tolerable loss before anything else: the price step divided by the contribution per unit after the step, which on the case is $10.00 divided by $74.00, or 13.51%, being 28 of 206 subscriptions. (3) Write down the single true reason the buyer will be told, and refuse to proceed until it fits in one sentence. (4) List the populations that do not follow the general rule — prepaid customers, customers under a fixed term, consumers, and any named legacy list — and decide each one explicitly. (5) Choose the notice period and the grandfathering period together, and price the grandfathering: on the case, six months of grandfathering costs $12,000.00 of the year's gain. (6) List every place a price is published — website, price list, quotation template, catalog, marketplace listing, partner material, invoice template — and assign each one an owner and a date, with the invoice last. (7) Telephone the largest accounts in descending order of revenue in the week before the letters. (8) Send one letter to every remaining customer on the same day, over the owner's signature, and change the advertising that same day. (9) Handle objections by exchange and not by exception, within a written authority that is narrower during the increase than outside it, recording every exception with its end date. (10) Ninety days after the effective date, count the customers lost and compare the count with the threshold from step two, and record both in the pricing file. Deliverable: a one-page increase plan holding the old and new prices, the step in dollars and in percent, the tolerable loss in customers, the notice and grandfathering dates, the populations treated separately, the publication checklist with owners and dates, and the measurement date. Verification criterion: ninety days after the effective date, the actual number of customers lost can be stated and compared with the tolerable loss computed before the increase, and no place where the price is published still shows the old price.

8.8 Lowering a Price, and When Not To

8.8.1 The arithmetic of a decrease is punishing, and it is permanent

Table 8.4 has already given the shape of the problem. Cutting the kit price by 10% on the case requires 21.62% more volume merely to stand still, and cutting the subscription by 10% requires 14.08% more. Those are not stretch targets; they are the definition of breaking even, and a business that cannot name where the extra volume would come from is proposing to work harder for less.

Two further properties make a decrease worse than its arithmetic. It is permanent in the buyer's mind, because Chapter 3 shows how a reference price forms: the price a buyer last paid becomes the normal price, and the lower price is the one that will be argued from at every renewal and repeated to every peer. And it is difficult to reverse, because reversing it is an increase, with everything this chapter has described attached to it, undertaken by a business that has just demonstrated that its price was negotiable.

8.8.2 The four situations in which lowering is right

There are four, and the fourth is not a price decrease at all. The first is a structural fall in the cost base that the business wants to convert into volume, where the competitor will convert it if the business does not; this is a strategic choice, it should be taken deliberately, and the break-even volume of Chapter 2 tells the business what it is committing to. The second is deliberate entry into a lower segment, which is properly a new offer with a fence around it rather than a lower price on the existing one — Book A6 (Designing a Value-Added Offer) designs the contents of that offer and Chapter 5 of this book builds the fence and sets the gap. The third is that the price was simply wrong: the measured value is below the price, the win rate has fallen, and the loss analysis says the losses are about price. That last condition matters, because Book E2 (Prospecting and Outbound Sales), Chapter 2, makes the distinction this book must not blur: a deal lost to a competitor with a fundamentally cheaper economic model is a signal about a segment that should be excluded, not a signal about the price. Chapter 9 gives the measurement that tells the two apart.

The fourth situation is a temporary, dated, fenced promotion, and it is included here only to be excluded. A promotion is not a price change; it is a price that returns. Chapter 6 treats it in full, with the authority, the fence and the end date. The reason it belongs in a chapter about change over time is that a promotion which is renewed indefinitely stops being a promotion. Once the discounted price is the price at which a substantial volume is actually sold, the former price is no longer a price at which the business ordinarily supplies the product, and a comparison to it becomes an ordinary-price claim that must satisfy the volume test or the time test — thresholds that come from the Competition Bureau's enforcement guidance, and not from the text of the Act (Competition Bureau Canada, 2022). A business that has been running the same promotion for two years has changed its price and is advertising a comparison it can no longer support.

8.8.3 How to lower a price without destroying the reference

Three practices make a decrease survivable. Change the offer at the same time, so that the buyer is comparing two different things rather than two prices for the same thing; removing something real from the lower-priced version is what makes the higher-priced version defensible afterwards. Use a fenced version — a smaller tier, a limited scope, a self-service channel — rather than a general cut, so that the customers who were paying the old price have a reason to continue. And if the decrease is temporary, state the end condition in the same sentence as the price, because a decrease without an end condition is a new price and will be treated as one.

Whatever is done, record it. The date the price changed, the price before, the price after and the reason belong in the pricing file, because the ordinary-price questions of Chapter 6 are answered from that record, the renewal conversations of Section 8.6 are prepared from it, and the measurement of Chapter 9 is impossible without it. A business that cannot say what it charged eighteen months ago cannot defend a comparison, cannot explain a difference between two customers, and cannot tell whether the last change worked.

IN LA MACHINE Changing a price is editing a field In La Machine, changing a price is exactly one action: opening the record and typing a different number. On Business → Session Packages (business/packages, and the same page reached as productivity/packages), on the subscription packages page (business/subscription-packages) and on Business → Store → Products (business/shop, business/shop-product-edit), the price field is edited and saved, and the margin the platform displays — profit per sale and percentage of margin for a package, the profit panel for a product — recalculates on screen against the cost field held beside the price. That pairing of price and cost is genuinely useful, and it is the reason a reader can see the new contribution as soon as the new price is typed. What is not here is everything else this chapter has described. There is no price history and no record of what the price was before, so the old price is gone the moment the new one is saved. There is no effective date: the change applies the instant it is saved. There is no grandfathering, no cohort, no legacy price and no link between an edited package and the clients who had already bought it. There is no proration, no scheduled change, no indexation and no index field, no contract term, no renewal date, no notice period and no price-change notification of any kind; the platform's promotion feature sends internal alerts and writes a simulated email line in a log, and it carries a whole-number percentage and two dates, which is a campaign device rather than a price change. The Control center section carries an Audit log, and this book makes no claim about whether price edits appear in it, because that was not read in the source. Routes verified in the route inventory on September 19, 2026, and behaviors read in the application's source at its commit of September 7, 2026 (La Machine, 2026).

Costly Mistakes

• Deciding whether to raise the price before computing how much loss the increase can absorb. The tolerable loss takes two minutes — the step divided by the contribution after the step — and it converts the decision from a bet into a threshold that can be measured against afterwards.
• Grandfathering without an end date. A grandfathering period stated as "for now" becomes a legacy price, which compounds with every later increase and eventually produces a business that cannot say what it charges without asking when the customer signed.
• Changing the invoice before the website. Supplying a product above the price advertised for it, during the period and in the market the advertisement covers, is reviewable conduct under section 74.05 of the Competition Act, and it happens by accident whenever the publication checklist is missing.
• Signing a multi-year price with no indexation and no cap. On the case, one step was worth $19,392.00 in a full year; a three-year freeze concedes that amount repeatedly, to a buyer who never asked for the certainty and would have paid for it.
• Writing "adjusted for inflation" instead of naming the series, the geography, the basis, the reference month and the fallback. An unnamed index converts an automatic adjustment into an annual negotiation, which is the opposite of what the clause was for.
• Offering an exception in the increase letter itself. An exception offered to a population that has not asked for one is granted to everybody, costs more than the increase earns, and teaches the whole base that the new price is a starting position.

Key Takeaways

• A price that is never edited is falling, because everything underneath it moves: Canadian consumer prices rose 6.8% on an annual average basis in 2022 and 3.0% in the twelve months to August 2026, while the case's variable cost, the value it delivers and its competitors' prices all moved too.
• There are six legitimate reasons to raise a price, and each requires evidence the business can produce: risen costs, an improved offer, higher measured value, a price that was wrong, a binding capacity constraint, and a repositioning.
• When capacity binds, price is the honest rationing device, because a queue rations by patience rather than by value, and the evidence to hold is operational: utilization, lead time, refused work and overtime.
• The tolerable loss is the step divided by the contribution after the step, which on the case is $10.00 divided by $74.00, or 13.51%, being 28 of 206 subscriptions — against 6 actually assumed to be lost.
• Increases are structurally safer than decreases: a 5% increase on the kit can absorb an 8.16% fall in volume, while a 10% cut on the same kit demands 21.62% more volume just to break even.
• Grandfathering is time bought with money, and the case's six months cost $12,000.00 of the year's $24,000.00 — a good purchase with a date on it, and a permanent loss without one.
• An unindexed multi-year price is a discount nobody negotiated, and the three honest alternatives are a fixed percentage, a named public index, and an index with a cap, which is a product the seller should charge for and should pair with a floor.
• The advertising changes first and the invoice changes last, because the sequence is what keeps an ordinary price increase from becoming reviewable conduct.

Your Action Plan

1. Date your prices. The deliverable is a one-line entry for every price you charge, giving the amount and the date it was last changed; it takes an hour with your price list and your invoices, and it is done when no price on the list has an unknown date.
2. Compute the tolerable loss for every offer. The deliverable is a table with, for each offer, a candidate increase, the contribution after that increase, the break-even percentage loss and the number of customers that percentage represents; it takes two hours with your margin model, and it is done when every offer carries a threshold you could compare a result against.
3. Write the one true sentence for the increase you are planning. The deliverable is a single sentence naming what changed, with the document behind it attached; it takes thirty minutes, and it is done when someone who does not work for you reads the sentence and does not ask a follow-up question.
4. List the populations that do not follow the general rule. The deliverable is a named list of prepaid customers, customers under a fixed term, consumers, and any legacy cohort, with the treatment decided for each; it takes an hour, and it is done when every customer in your base belongs to exactly one line of the list.
5. Build the publication checklist. The deliverable is a list of every place a price of yours appears — website, price list, quotation template, catalog, marketplace listing, partner material, invoice template — with an owner and a date for each, the invoice last; it takes an hour, and it is done when a colleague can work through it without asking you where else the price is shown.
6. Put an indexation mechanism in every agreement longer than twelve months. The deliverable is a standard clause naming the publisher, the series, the geography, the basis, the reference month, the rounding, the floor and the fallback, with a cap if you choose to sell one; it takes two hours with Book H1 in front of you, and it is done when someone in your office could compute next year's price from the clause alone.
7. Schedule the measurement before you send the letter. The deliverable is a calendar entry ninety days after the effective date, holding the count of customers before, the count kept, the count lost, the tolerable loss you computed in advance, and the revenue effect; it takes fifteen minutes to create, and it is done when the comparison has been written into the pricing file.

In La Machine

Every route in this section was verified in the route inventory on September 19, 2026, and the behaviors were read in the application's source at its commit of September 7, 2026 (La Machine, 2026); Book N8 will cover these pages one by one, and this section says only what a price change can and cannot be expressed with today.

• Editing a price is a single field. On Business → Session Packages (business/packages, productivity/packages), on business/subscription-packages and on Business → Store → Products (business/shop, business/shop-product-edit), the price is typed and saved, and the profit per sale and the percentage of margin recalculate on screen against the cost field stored beside the price.
• The cost field is what makes the new price legible. Because price and cost sit together on the same record, the contribution after an increase is visible as soon as the new number is entered, which is the one part of this chapter the platform genuinely supports.
• A subscription package carries a billing period that is monthly or annual and nothing else, so a prepaid year and a monthly subscription can both be represented, but no term, no renewal date and no notice period can be.
• A promotion carries a whole-number percentage, a status set by hand, a start date and an end date (business/promotions, and the same page reached as offers/promotions), and sending it writes internal alerts and a simulated email line in a log; the dates bound a campaign, and they are the closest thing on the platform to a dated price.
• "Revenues & Expenses" (business/subscriptions, reached as Account → Profile → Revenues & Expenses) is the only view that places a revenue beside its cost, which is where the result of an increase would be read if the platform recorded one.
• What is not here. There is no price history and no record of any earlier price, no effective date, no scheduled change, no grandfathering, no cohort, no legacy price and no link between an edited package and the clients who bought it; there is no proration, no indexation and no index field, no contract term, no renewal date, no notice period and no price-change notification; the checkout is a demonstration (/checkout-demo) and /client-subscribe is disabled, so nothing is billed at either the old price or the new one; and the platform checks no price presentation against the Competition Act or the Consumer Protection Act, which Book B1, Chapter 6, states already and this chapter confirms. The Control center section carries an Audit log, and nothing here relies on it, because what it records was not read in the source for this book.

Self-Check

1. Why is a price that has not been edited for three years described in this chapter as a price that has fallen, and what evidence does the chapter offer for the claim?
2. What are the six legitimate reasons to raise a price, and which of them is a response to a constraint on delivery rather than on cost?
3. How is the tolerable loss computed, what is it on the case, and how does it compare with the loss the model assumes?
4. Why is a decrease harder than an increase of the same size, and what do the two kit figures in Table 8.4 show?
5. What six things must an indexation clause name for it to be applied without an argument, and why does "adjusted for inflation" fail?

Answers. (1) Because everything the price is made of moves while the price waits to be edited: Canadian consumer prices rose 6.8% on an annual average basis in 2022 and 3.0% in the twelve months to August 2026, the case's variable cost of $428,580.00 sits against revenue of $1,130,388.00 for a gross margin of 62.09%, the value delivered has been measured at $5,668.00 a year against $1,908.00 charged, and the company's own new customers already pay $89.00 while 206 existing buildings pay $79.00 (Sections 8.1.1 to 8.1.3 and Table 8.1). (2) Risen input costs, an improved offer, measured value higher than assumed, a price that was set too low, a binding capacity constraint, and a repositioning; the capacity constraint is the one that concerns delivery rather than cost, and it is rationed honestly by price rather than by queue (Sections 8.2.1 and 8.2.2, Table 8.2). (3) It is the price step divided by the contribution per unit after the step, which is $10.00 divided by $74.00, or 13.51%, being 28 of 206 subscriptions; the model assumes 6 are lost, so the assumption would have to be nearly five times larger before the increase stopped paying (Sections 8.3.3 and 8.4.1 to 8.4.2). (4) Because a decrease shrinks the contribution per unit while the price change stays the same, so the required volume gain always exceeds the tolerable loss of an equivalent increase: the kit can absorb a 8.16% fall in volume on a 5% increase, while a 10% cut on the same kit demands 21.62% more volume merely to break even (Sections 8.4.1 and 8.4.3, Table 8.4). (5) The publisher and series, the geography, the basis, the reference period, the rounding and floor, and the fallback if the series is discontinued or rebased; "adjusted for inflation" fails because several series, several geographies and several bases all answer to that word, so two reasonable people produce two different numbers and the automatic adjustment becomes an annual negotiation (Section 8.6.3).

Chapter References

ChartMogul. (2025, December 10). The SaaS retention report: The AI churn wave. Retrieved September 19, 2026, from https://chartmogul.com/reports/saas-retention-the-ai-churn-wave/ (annualized 2025 rates from approximately 3,500 vendor websites categorized with the assistance of artificial intelligence; a non-probability sample)

Competition Act, RSC 1985, c C-34, ss. 74.01, 74.05 and 74.1 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Competition Bureau Canada. (2022, June 23). Ordinary price claims. Government of Canada. Retrieved September 19, 2026, from https://competition-bureau.canada.ca/how-we-foster-competition/education-and-outreach/publications/ordinary-price-claims

Consumer Protection Act, CQLR c P-40.1, s. 224 (Québec). LégisQuébec, consolidation of April 7, 2026, as read for Book E1 on September 8, 2026; not read again, LégisQuébec returning an access error on September 19, 2026.

Faustino, S. (2025, May 18). SaaS billing report: Billing is the invisible force behind your growth. ChartMogul. Retrieved September 19, 2026, from https://chartmogul.com/reports/saas-billing-report/ (aggregated billing data from more than 2,500 subscription software companies for full-year 2024; a non-probability sample of one vendor's own customers, excluding companies below $300,000 of annual recurring revenue)

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Statistics Canada. (2023, January 17). Consumer Price Index: Annual review, 2022. The Daily. Retrieved September 19, 2026, from https://www150.statcan.gc.ca/n1/daily-quotidien/230117/dq230117b-eng.htm

Statistics Canada. (2026a, June 18). Analysis on small businesses in Canada, second quarter of 2026 (Catalogue no. 11-621-M). Retrieved September 19, 2026, from https://www150.statcan.gc.ca/n1/pub/11-621-m/11-621-m2026008-eng.htm

Statistics Canada. (2026b, September 14). Consumer Price Index, August 2026. The Daily. Retrieved September 19, 2026, from https://www150.statcan.gc.ca/n1/daily-quotidien/260914/dq260914a-eng.htm

Statistics Canada. (2026c). Table 18-10-0004-01: Consumer Price Index, monthly, not seasonally adjusted [Data table]. Retrieved September 19, 2026, from https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1810000401

Chapter 9 — Measuring Pricing Performance

In the twelve months that this book has followed it, Hydrosentinelle inc. has changed almost everything about the way it prices. It replaced a single subscription with three published tiers, wrote a margin floor and an authority ladder, rebuilt its quotation so that the total payable appears before anything else, and moved two hundred and six existing customers from one monthly price to another. Each of those decisions was defended with an argument, and each argument was plausible. None of them has yet been checked. At the end of that year the president, Laurence Bergeron, asked the cofounder responsible for sales and operations, Karim Haddad, a question that sounds simple and is not: did any of it work? The only document in the company capable of answering was a spreadsheet with 96 rows, one for each quotation sent during the year, carrying the date, the buyer, the list value, the discount applied and the outcome. This chapter is what that spreadsheet says. It says that 34 quotations were won, a win rate of 35.42 percent; that of the $163,200.00 of list value won, $148,848.00 arrived in the company's pocket, a realization of 91.21 percent; and that the most profitable behavior of the year was not the one anybody would have guessed. It also says something the company did not want to hear about testing, which Section 9.7 states without softening it.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Build a pricing dashboard of six numbers that a business of any size can produce from records it already keeps, and know what each number answers.
• Compute your price realization from list value to invoice value to pocket value, and read the gap as the cost of a year of concessions rather than as bad luck.
• Read your win rate correctly, including the denominator problem that makes most reported win rates meaningless, and recognize a win rate that is too high.
• Read a year of discounts by band and find the band where discounting stops paying, which no average and no single curve will show you.
• Measure willingness to pay without asking anyone what they would pay, using past purchases, field tests, conjoint analysis and the price sensitivity meter, each with its limits stated.
• Decide whether your business can test a price at all, by computing the sample an honest test would need before you run one, and choose the method that fits the volume you actually have.

Why It Matters

Everything in the eight chapters before this one is a hypothesis. A value estimate is a hypothesis about what a buyer gains, a tier structure is a hypothesis about how buyers sort themselves, a floor is a hypothesis about what the market will bear, and a price increase is a hypothesis about who will stay. Measurement is the only operation that converts any of them into knowledge, and it is the one most small businesses skip, not out of laziness but because nobody ever told them which six numbers matter and where those numbers come from. The cost of skipping it is not an error in one decision; it is that every pricing decision in the business remains a matter of opinion forever, and opinions are settled by whoever speaks with most confidence rather than by what happened.

For a solo business, this chapter matters because the owner is both the person who granted every discount and the person who must now judge them, and a number is the only referee available. Ten minutes a month spent adding a row to a quotation log produces, after a year, the single most valuable pricing document the business owns. For a small team, it matters because measurement replaces argument: when a seller says the deals cannot be won without a deeper discount, the reading of Section 9.5 answers on the evidence rather than on seniority, and the answer on the case is that the deepest band bought wins the company could not afford. For an established business, it matters because the numbers can be cut by segment, product, channel and seller, and because those cuts routinely reveal that a fifth of the customer base is served at a loss while another fifth pays list without complaint.

There is a harder reason, and it governs the second half of the chapter. The measurement methods that pricing literature describes were designed for businesses with thousands of transactions, and most of them cannot be executed by a business with 96 quotations a year. This chapter computes exactly how far short the volume falls — detecting a 5 point improvement in the win rate would take 1,477 quotations in each arm of a test, 2,954 in all, about 30.8 years — and then does the useful thing, which is to set out what does work at that volume. A book that described the ideal method and left the reader to discover its impossibility would be worse than useless, because the reader would conclude that measurement is for other people.

In This Chapter

• 9.1 The Six Numbers a Pricing Dashboard Needs
• 9.2 Price Realization
• 9.3 Win Rate, and What a High One Says
• 9.4 Margin by Segment, Product, Channel and Seller
• 9.5 The Distribution of Discounts
• 9.6 Measuring Willingness to Pay Without Asking the Question
• 9.7 Price Testing
• 9.8 The Quarterly Pricing Review

9.1 The Six Numbers a Pricing Dashboard Needs

9.1.1 What a pricing dashboard is for, and what it is not

A pricing dashboard is not a financial statement and it is not a sales report. The financial statement says what the business earned; the sales report says who sold what; the pricing dashboard says whether the prices the business publishes are the prices it actually gets, and what it gave away to get them. Those are different questions, and none of the three documents answers the other two. A business can have an accountant, a clean set of statements and a monthly sales meeting, and still have no idea that a tenth of its list margin leaves the building every year in concessions that nobody records as a price decision.

The dashboard has three properties that distinguish it from the reports a business already produces. It measures against list, so that every number is expressed as a share of what the business intended to charge rather than as an absolute that rises and falls with volume. It measures the quotations that lost as well as the ones that won, because a price is a decision about a population of opportunities and the losses carry at least as much information as the wins. And it measures dispersion, not only averages, because the average discount is the one number in pricing that is almost always uninformative, for reasons Section 9.5 demonstrates on the case.

It also has to be small. A dashboard of thirty indicators is a report nobody reads, and a business of the size this book addresses will maintain six numbers or none. The six below were chosen on three criteria: each answers a question an owner actually asks, each can be computed from records a business already has or can start keeping in an afternoon, and each, when it moves, points to a specific chapter of this book for the remedy.

DEFINITION Price realization, win rate, discount dispersion and margin per quotation sent Price realization is the share of list value that the business actually keeps, measured by dividing the pocket value of what was sold by the list value of the same sales; measured at the invoice it is called invoice realization, and measured after off-invoice concessions it is pocket realization, the two differing by everything that never appears on an invoice. Win rate is the number of quotations won divided by the number of quotations sent in the same period, and it is meaningful only when the denominator is a complete and honestly maintained count of everything sent, including the quotations that expired without an answer. Discount dispersion is the spread of realized prices around their average for one and the same product, usually read as a distribution across discount bands rather than as a single statistic; Marn and Rosiello (1992) named this reading the pocket price band, and the publisher's description of their article reports that a spread of more than 60 percent between the highest and the lowest price actually realized for the same product is common among manufacturers. Margin per quotation sent is the total margin a band or a segment produced divided by the number of quotations sent into it, and it is the measurement that makes discount bands comparable, because it charges each band for the effort it consumed as well as crediting it with the money it earned. In plain terms: realization says what you keep, win rate says how often you win, dispersion says how unequal your prices really are, and margin per quotation sent says which behavior actually pays.

9.1.2 The six numbers

The table that follows is the dashboard itself. Its five columns give, for each number, the question it answers, the way it is computed, the value it takes on the case at the end of year 3, and how often it should be read. Read the last column with care: three of the six are quarterly rather than monthly, because a monthly reading of a business with eight quotations a month measures noise, and a number read too often trains everyone to react to nothing.

Table 9.1 — The six numbers a pricing dashboard needs

Two of the six values in that table are assumptions rather than measurements, and the table says so rather than hiding it. The mix of 55, 35 and 10 percent across the three tiers is what the model assumes and what Chapter 5 used to compute an average revenue per building of $112.00; the company has not yet counted how its new buildings actually divided. The attach rate of the sensor pack has never been measured at all, which Chapter 5 admitted when it computed the pack's contribution of $270.00 and observed that packs attached to all 180 of the year's kit installations would add $48,600.00, or 6.92 percent of the year's gross margin of $701,808.00. A dashboard whose cells contain assumptions is not a failure, provided the cells are labeled. A dashboard whose cells contain assumptions presented as measurements is worse than no dashboard, because it manufactures confidence.

9.1.3 Where the numbers come from when nothing is instrumented

The objection an owner raises at this point is that the business has no system capable of producing any of this. The objection is usually true and it is not an obstacle, because five of the six numbers can be reconstructed from documents that already exist.

• The quotation log is the one artefact that has to be created, and it is a single sheet with one row per quotation sent, carrying the date, the buyer, the list value, the discount applied, the option recommended, the option accepted and the outcome. Chapter 7 ends by asking the reader to start exactly this sheet, and it is the input to three of the six numbers.
• The invoices already issued give the invoice value of everything won, and subtracting them from the list value of the same orders gives invoice realization without any new record-keeping at all.
• The accounting records give the cost of serving one unit, which Chapter 2 insists must be agreed with whoever is responsible for the numbers rather than remembered, and which turns a price into a contribution in dollars.
• The order files give the off-invoice concessions — free goods, unbilled work, extended payment terms, absorbed freight — and these are the lines that separate invoice realization from pocket realization, as Chapter 6 showed on a single order.
• The customer list, with start and end dates, gives retention, and a list of names with two dates beside each is enough to compute it for a business of this size.

Reconstructing a year backwards from these sources takes a day or two of unpleasant work, and it is worth doing once, because the alternative is to wait a year for the log to fill before learning anything. After that first reconstruction the log carries the burden, at a cost of about a minute per quotation.

WHO DOES THIS Who reads the pricing numbers, and how often In a solo business the owner is the only candidate, and the practical discipline is to separate the two roles that the owner is playing: the person who granted the discounts is a poor judge of them, so the reading is done on a fixed date, from the log, without reference to memory, and the conclusions are written down before any explanation is offered. Thirty minutes a quarter is enough. In a small team, the reading belongs to the person accountable for margin rather than to the person accountable for revenue, because the two roles are in genuine tension and the dashboard is precisely the instrument that resolves it; at Hydrosentinelle the log is maintained by Karim Haddad, who leads sales and operations and prepares every quotation, and it is read with Laurence Bergeron, the president, who approves anything below the floor. The seller maintains the record and does not grade it. In an established business three roles appear: someone owns the definitions, so that win rate means the same thing in every report; someone owns the extract, so that the numbers are produced the same way each quarter and can be compared; and someone owns the decision, which is the person who may change a price. The failure mode at every size is the same and it has a name: a number produced by the person whose behavior it measures, read by nobody, in a meeting with no decision on its agenda. Book E8 (Sales Management: Pipeline, Forecasting, Coaching and Compensation) governs how a seller is managed against these numbers; this chapter governs only the numbers.

IN LA MACHINE Revenues and expenses, and nothing else Of everything this chapter measures, La Machine holds one piece, and it is a genuine one. The page reached at Account → Profile → Revenues & Expenses (business/subscriptions) reads the subscription transactions, the package purchases and the product purchases recorded in the platform and writes each sale as a revenue line with its cost lines beneath it; it is the only view in the product that places a revenue next to what it cost to produce, and it is therefore the only place where anything resembling a realized margin can be seen. Two qualifications belong with that statement. The sample data it ships with are generated, with Canadian sales taxes of 5 percent and 9.975 percent written into the generator, so a reader looking at an unfamiliar screen should confirm which rows are real. And it is a list of transactions rather than an analysis: it totals, but it computes no realization, no win rate and no distribution. The margin fields elsewhere in the product are equally real and equally narrow — Business → Session Packages (business/packages) and the subscription packages page (business/subscription-packages) each hold a price and a cost and display profit per sale and a margin percentage on screen, and Business → Store → Products (business/shop) holds the product cost, the shipping cost and the shipping fees with a profit panel. Those screens answer the fourth number of Table 9.1, one offer at a time, and none of the other five.

9.2 Price Realization

9.2.1 Three values for the same year

Chapter 6 followed a single order from its list price of $4,800.00 down to a pocket price of $3,974.12 and showed that 30.59 percent of the list margin disappeared in six lines of which only one was negotiated as a discount. Price realization is that same reading, performed on a whole year instead of one order, and it produces a number an owner can hold in mind: the share of what the business publishes that the business actually keeps.

Three values are needed, and the discipline is to compute all three rather than the two that are easy.

The list value of what was won is the number of units sold multiplied by the published price, before any reduction. It is the counterfactual: what the year would have produced if every buyer had paid the price on the price list. Businesses rarely compute it, because no document contains it, and its absence is the reason realization goes unmeasured for years.

The invoice value is the sum of what was actually billed. It captures every reduction that a seller wrote down: volume bands, negotiated discounts, promotional prices. It is easy to obtain and it flatters the business, because the reductions a seller writes on an invoice are the honest half of the story.

The pocket value is the invoice value less everything granted that never reached an invoice. Free goods, unbilled work, absorbed freight, the cost of payment terms longer than standard, a service credit issued to settle a complaint: each is a transfer of margin to the customer, and none appears as a discount anywhere in the accounts. Chapter 6 priced each of these on one order; at the level of a year they are estimated as a rate, which is what the case does at 3 percent of list value won.

9.2.2 The case's year, measured

The table that follows performs the calculation on the year Hydrosentinelle has just finished. Its four columns give the step, the amount, that amount as a share of the list value won, and what the step records. Read the bottom row first: of every dollar the company published, it kept about ninety-one cents, and the nine cents it lost did not go anywhere it could name without this calculation.

Table 9.2 — From list value to pocket value on one year of won deals

The off-invoice line deserves one sentence of method, because it is the line most readers will be tempted to leave out. The case estimates it at 3 percent of list value won rather than adding up individual instances, on the ground that a business of this size will not capture every free sensor and every extra installation hour at the moment it happens. An estimate stated as an estimate is the right answer here; a zero is the wrong one, because Chapter 6 measured the off-invoice lines of a single order at $312.28, which was 6.51 percent of that order's list price, more than twice the rate assumed here. The rate should be revised the first year the business captures the concessions properly, and it will almost certainly be revised upwards.

DO THE MATH Price realization across one year of won deals Fictional example; the discount bands, the off-invoice rate and the volumes are illustrative assumptions of the model. Start from what the year produced. Hydrosentinelle won 34 kit orders at a published price of $4,800.00, so the list value won was 34 times $4,800.00, which is $163,200.00. The invoices for those same orders totalled $153,744.00, because eleven of them carried a discount averaging three percent, ten carried eight percent and six carried fourteen percent, while seven carried none. Invoice realization is therefore $153,744.00 divided by $163,200.00, which is 94.21 percent, and the invoice discount averaged 5.79 percent across the won orders. Now subtract what never reached an invoice, estimated at 3 percent of the list value won, which is $4,896.00. Pocket value is $153,744.00 less $4,896.00, which is $148,848.00, and pocket realization is $148,848.00 divided by $163,200.00, which is 91.21 percent. The margin arithmetic is harsher than the price arithmetic, and this is the point of the exercise. At list, 34 kits would have earned 34 times $2,700.00, which is $91,800.00 of margin. The invoiced margin was $82,344.00, which is 89.70 percent of that, and the pocket margin was $82,344.00 less $4,896.00, which is $77,448.00, or 84.37 percent. The company gave up 8.79 percent of its price and 15.63 percent of its margin, and the difference between those two percentages is the whole reason a margin rate of 56.25 percent must never be used to reason about a discount.

9.2.3 What a realization number proves, and what it does not

A realization figure is a diagnosis, not a verdict, and three cautions keep it honest.

The first is that there is no benchmark. A realization of 91.21 percent is neither good nor bad in the abstract, and any consultant who says otherwise is selling a number they cannot source. Realization is meaningful against two things only: your own figure last year, and the figure your policy intended. Hydrosentinelle's policy, written in Chapter 6, allows a representative 6 percent and the head of sales 12 percent; a pocket realization of 91.21 percent means the average concession across the year, counting everything, was 8.79 percent, which sits inside the policy in aggregate while telling you nothing about whether any individual deal stayed inside it.

The second is that realization falls for good reasons as well as bad ones. A business that deliberately introduced a volume band, as Chapter 5 did with its 6 percent band at five buildings and 12 percent band at twenty, has chosen to give up realization in exchange for larger orders. The measurement does not distinguish a structural discount the business designed from a negotiated discount a seller conceded, which is why the log must record which band or which negotiation produced each reduction. A realization number without that breakdown will be blamed on the sellers, and half the time the sellers will be innocent.

The third is that realization is measured on what was won, and therefore says nothing about what was lost. A business can achieve 100 percent realization by refusing every negotiation and losing every deal. That is why realization is read beside the win rate and never alone, and it is why the two numbers together are still not enough, which brings the chapter to the reading that makes them useful.

9.3 Win Rate, and What a High One Says

9.3.1 The definition, and the denominator that ruins it

The win rate is quotations won divided by quotations sent in the same period. The arithmetic is trivial and the measurement is routinely wrong, because the denominator is the part nobody maintains. Four failures are common, and each one inflates the number.

• Quotations that were never formally sent are excluded, so the verbal price given on a site visit and refused on the spot never enters the count, and the business measures only the opportunities it considered promising enough to document.
• Quotations that expired without an answer are dropped, on the reasoning that the buyer never said no. A quotation that expires is a loss with better manners, and Chapter 7 asks for expiry to be recorded as an event precisely so that this denominator stays honest.
• The period is chosen after the fact, so a quarter with a large win is compared with a year containing several losses, and the comparison means nothing.
• Revisions are counted as new quotations, which doubles the denominator for the deals that took most work, or, in the other direction, the original is deleted and only the accepted revision survives, which deletes the evidence of what was first asked.

The remedy is a definition written once and applied mechanically: one row per quotation sent to a buyer, revisions linked to the original rather than counted again, outcomes recorded as won, lost or expired, and no row ever deleted. Hydrosentinelle's year contains 96 rows on that definition, of which 34 were won, giving 35.42 percent. A business that changes the definition mid-year has lost the ability to compare itself with itself, which is the only comparison available.

9.3.2 What a high win rate says

Owners are pleased by a high win rate and should not be. A win rate is the product of two things: how good the business is at choosing which opportunities to pursue, and how generous its price is relative to what buyers will accept. Improving the first is the work of Book E2 (Prospecting and Outbound Sales) and of the qualification discipline in Book E1 (The Sales Process: From Prospect to Signed Contract). The second is a pricing decision, and it moves the win rate in a way that is easy to mistake for sales performance.

The reasoning is arithmetic. If nearly every quotation is accepted, and accepted quickly, and without the buyer contesting the price, then the price is below what that population of buyers would have paid, and the business is leaving money on every one of those transactions. The exception is real and must be stated: a business with a genuinely differentiated offer, a qualification process that rejects most enquiries before a quotation is prepared, and a price that buyers argue about before accepting can sustain a high win rate honestly. The diagnostic is not the number alone but the number together with the behavior: if nobody negotiates, the price is too low, whatever the win rate.

KEY CONCEPT A win rate above eighty percent is a price that is too low, not a sales team that is good A business that wins four quotations out of five is usually being congratulated when it should be investigating. There are only three explanations for a win rate that high, and two of them are expensive. The first is that the business qualifies ferociously and only quotes the opportunities it has already effectively won, in which case the win rate is measuring the qualification process and the real question is how many opportunities are being discarded that could have been sold at a price nobody tested. The second is that the offer is genuinely without substitute in its market, which is possible, rare, and temporary. The third, which is the usual one, is that the price is below the point where buyers hesitate, and the evidence for it is not the win rate but the silence: nobody asks for a discount, nobody compares, nobody takes the quotation away to think about it. A price that no buyer ever contests has not been tested by the market. The practical response is not a large increase but a bounded experiment of the kind Section 9.7 describes: raise the price for one segment, or one channel, or new customers only, by an amount large enough to matter, and watch the win rate. If it does not move, the old price was a gift; if it falls by less than the break-even volume change that Chapter 2 computes, the increase paid for itself. The reverse reading matters too, and it is the one that sends a reader to Section 9.5: a win rate that climbs as discounts deepen is not evidence that discounting works, because the question is never whether a discount buys wins — it always does — but whether it buys enough of them.

9.3.3 A price problem or a targeting problem

A business that loses on price concludes that its price is wrong. Sometimes that conclusion is correct, and sometimes the loss is telling it something entirely different. Book E2 (Prospecting and Outbound Sales), Chapter 2, states the principle this chapter has to operationalize: a deal lost to a competitor with a fundamentally cheaper business model is information about a segment the business should exclude, not information about its pricing. A business that responds to that kind of loss by cutting its price will lose its margin and still lose the deal, because it cannot reach a cost structure it does not have.

Distinguishing the two takes a loss reason recorded at the moment of the loss, in the seller's words and in a small fixed vocabulary, plus four readings of the log. The table that follows gives the four readings. Its four columns state what you observe, what that observation suggests if the problem is the price, what it suggests if the problem is the targeting, and which action to take first. The rule that organizes it is simple enough to remember: a price problem shows up as losses to comparable competitors at small price gaps, and a targeting problem shows up as losses to a structurally different offer at gaps no discount could close.

Table 9.3 — Telling a price problem from a targeting problem

Two consequences follow, and both are uncomfortable. The first is that a segment identified as a targeting problem should be excluded, not discounted: the business stops quoting it, which raises the win rate, raises the average margin and frees the selling time that was being consumed by opportunities it could not profitably win. The second is that this reading requires the loss reasons to be recorded by the person who lost the deal, who has an interest in recording "price" for every loss, since "price" is the only loss reason that blames the company rather than the seller. A log in which every loss reason is "price" is not evidence about pricing; it is evidence that the field is unusable, and the fix is a short closed list of reasons, filled in within a day, with the competing price and the winner named.

9.4 Margin by Segment, Product, Channel and Seller

9.4.1 Contribution in dollars, never margin rate

Chapter 5 established the rule this section applies: offers are ranked by the contribution they produce in dollars, never by their margin rate, because a rate has no size. The same rule governs every cut in this section. A segment with a 70 percent margin rate that generates $8,000.00 of contribution a year is worth less than a segment with a 45 percent rate that generates $60,000.00, and a business that manages by rate will spend its effort on the first and wonder why profit does not move.

The cut that matters is therefore always the same shape: for each slice of the business, the number of units, the pocket revenue, the cost of serving those units, the contribution in dollars, and — if quotations are tracked, which they should be — the number of quotations sent into that slice. The last column is what makes slices comparable in effort as well as in money, and Section 9.5 shows what it reveals when it is added.

9.4.2 The four cuts, and the fields each one requires

Four cuts answer four different questions, and each requires a field to exist on the record before the cut can be made. A business that has never made these cuts almost always finds that one field is missing, and the whole discipline of this section is to add that field now so that next year's reading is possible.

• By segment answers whether a kind of buyer is worth serving. The field required is a segment tag on every customer and every quotation, chosen from a short list fixed in advance — for the case, a single-building owner, a small portfolio manager, a large portfolio manager and an institution. Hydrosentinelle does not have this field, which is the subject of the case note below.
• By product or tier answers whether the structure built in Chapter 5 is behaving as designed. The field required is the tier or product on every line, which the business already has. On the case, the kit contributes $2,700.00 a unit, the old single subscription $64.00 a month, and the three tiers $74.00, $98.00 and $141.00 a month respectively against costs of service of $15.00, $31.00 and $38.00.
• By channel answers whether a route to market pays for itself after its own costs. The field required is a channel tag, and the arithmetic must charge each channel with what it costs: a referral fee, a partner commission, the extra discount a channel demands. A channel measured on revenue always looks good.
• By seller answers whether the authority ladder of Chapter 6 is being used consistently. The field required is the person who prepared each quotation. This cut is the most politically charged and the most useful, because discount behavior varies far more between individuals than between segments, and the variation is invisible until it is measured.

A caution belongs with all four. These cuts are built from customer records, and using customer records for analysis carries obligations about consent and about what may be kept and for how long; Book B1 (Marketing Fundamentals for Business Owners), Chapter 10, treats those obligations, and this chapter assumes the analysis is performed on records the business already holds lawfully for billing.

9.4.3 Mix, attach rate and the option nobody chooses

Three measurements read the structure rather than the money, and all three come from the quotation log rather than from the accounts.

The mix is the share of new customers landing on each tier. Chapter 5 computed an average revenue per building of $112.00 a month from an assumed division of 55 percent to Essential, 35 percent to Plus and 10 percent to Portfolio, and said plainly that the division was an assumption of the model rather than a measurement of the market. Measuring it requires nothing more than counting the tier on each new subscription and dividing. If the realized mix is poorer than the assumption — more Essential, less Plus — the average revenue falls and the uplift the structure was built to produce shrinks, and the business will see it in the average long before it sees it in the profit.

The attach rate is the share of primary sales carrying an add-on, computed as add-ons sold divided by primary units sold. On the case it is the sensor pack, priced at $390.00 against a cost of $120.00 and contributing $270.00, and the company has never measured it. Counting packs sold and dividing by the 180 kits installed is a five-minute calculation that the business has not yet performed, which is an honest admission and an extremely common one.

The option mix is the measurement that answers the symptom Book E1 (The Sales Process: From Prospect to Signed Contract), Chapter 14, sends to this book: options that nobody chooses. Chapter 5 gave the four possible causes — an invisible fence, a wrong gap, a segment that does not exist, or an option that is never actually presented — and said that the cheapest test comes first because it disqualifies the other three. Here it is. Take the year of quotations and count two distributions rather than one: how often each option was quoted, and how often each option was accepted. An option quoted in most documents and accepted in none is mis-spaced or undifferentiated, and the remedy is in the gap or the fence. An option that appears in a small share of documents is not being offered at all, usually because the person preparing the quotation stopped including it after an early refusal, and the remedy is the template rather than the price. The two readings cost an hour of counting and they point at different chapters, which is why counting them before changing anything is the whole of the advice.

9.4.4 Retention and lifetime, measured rather than assumed

A price does not only win a customer; it selects a customer, and the customers a low price wins are not the customers a high price wins. The measurement that captures this is retention, computed as customers lost in a period divided by customers exposed to loss in that period, and read separately for each price level, each tier and each billing period.

The case assumes an annual churn of 10 percent, which produced 30 lost subscriptions in year 3 against an average base of 281. An assumed churn rate implies a customer life of one divided by the rate, which is ten years here, and at a subscription margin of $64.00 a month that implies a lifetime contribution of $7,680.00. That number should be treated with suspicion, and the reason is worth stating because it applies to every young business: a company in its third year has not observed a ten-year life and cannot have measured one. What it has is an assumed rate extrapolated to a horizon several times longer than its own history. The honest practice is to report what has actually been observed — how many of the customers who started two years ago are still here — and to label the extrapolation as an extrapolation whenever a lifetime figure is used to justify an acquisition cost or a promotional discount.

The same caution applies to retention figures borrowed from published reports, and Chapters 4 and 6 have already used two of them. A billing report published on May 18, 2025, covering the full year 2024 and built from aggregated billing data of more than 2,500 subscription software companies, reports that at low revenue per account customer retention was 62 percent on annual plans against 41 percent on monthly plans (Faustino, 2025); a retention report published on December 10, 2025, covering 2025 and built from approximately 3,500 vendor websites categorized with the assistance of artificial intelligence, reports that among artificial-intelligence-native products those priced above $250 a month showed 70 percent gross retention against 23 percent for those below $50 (ChartMogul, 2025). Both are non-probability samples of one vendor's own data, without peer review, drawn from subscription software rather than from building services. They are a reason to form the hypothesis that a low price attracts customers who leave, and they are not a measurement of your business. The measurement of your business is the two dates beside each customer's name.

TOOLS OF THE TRADE What an analytics tool must be able to slice Whatever a business uses to read these numbers, the requirement is not charting and it is not artificial intelligence; it is the ability to slice one table of transactions by the fields that matter and to export the result. It must hold one row per quotation sent, not only per order won, because every measurement in this chapter that involves a denominator dies without the losses. It must carry, on each row, the list value as well as the invoiced value, since realization cannot be computed from invoices alone. It must accept off-invoice concessions as amounts attached to the order rather than as a note, so that pocket value is computable rather than estimated forever. It must carry the discount band, the segment, the channel, the tier or option quoted and accepted and the person who prepared the document, because those five fields are the four cuts of this section plus the reading of Section 9.5. It must support a date range that the reader chooses, so that a quarter can be compared with the same quarter of the previous year rather than with whatever the tool decided to show. It must compute margin per quotation sent, which almost no standard report offers and which is the single most useful derived figure in this chapter. It must keep history, so that a price list in force on a past date can be recovered and a realization recomputed against the price that actually applied. And it must export in an open format, because the analysis that matters will eventually be done in a spreadsheet by someone who wants to try a cut the tool does not offer. A spreadsheet with those columns satisfies every requirement on this list for a business of the size this book addresses, and Chapter 12 maps the same functions against market tools and against La Machine.

9.5 The Distribution of Discounts

9.5.1 The average discount is the least informative number in pricing

A business that measures its discounting at all usually measures the average discount, and the average discount is close to useless. Two businesses with the same average of 6 percent can be in completely different situations: one grants 6 percent to nearly everyone as a structural band, and the other grants nothing to two-thirds of its buyers and 18 percent to the rest. The first has a price list that is 6 percent too high and should be rewritten; the second has a policy problem and a fairness problem, since buyers eventually discover what other buyers paid.

What carries the information is the distribution: how many transactions sat at each level of discount, and what each level earned. Marn and Rosiello (1992) made this reading famous under the name of the pocket price band, and the publisher's description of their article reports that a spread of more than 60 percent between the highest and the lowest realized price for the same product is common among manufacturers. The number to notice in that sentence is not the sixty; it is the fact that the businesses concerned did not know it until someone drew the distribution.

For a small business the distribution is drawn in bands rather than as a continuous curve, because there are not enough transactions for anything finer. Four bands are enough: no discount, a small one, a moderate one, and everything above the point where approval becomes necessary. The bands should be chosen to align with the authority ladder, so that the reading answers a governance question as well as an economic one, and on the case they align with the 6 percent a representative may grant and the 12 percent the head of sales may grant.

9.5.2 The case's year, band by band

The table that follows is the reading that organizes this chapter, and it is the one measurement in this book that overturns an intuition rather than confirming one. Its six columns give the discount band, the quotations sent into it, the quotations won from it, the resulting win rate, the margin earned for every quotation sent, and the total margin the band produced. Read the fourth and fifth columns against each other: the fourth rises from band to band without exception, and the fifth does not.

Table 9.4 — One year of quotations by discount band

The win rate column settles one argument permanently: discounting works. A quotation carrying a discount above ten percent was won 46.15 percent of the time against 22.58 percent for a quotation carrying none, which is more than double. Any owner who has told a sales team that discounts do not buy deals is contradicted by this column, and should say so out loud before reading the next one.

The margin per quotation sent column settles the argument that actually matters. It peaks at $1,004.14 in the 1 to 5 percent band and then falls, to $965.00 and then to $936.00, even though the win rate keeps climbing. The deepest band won the most often and earned the least per unit of selling effort. That is the finding, and it is invisible in every other way of looking at the same data: the total margin column makes the 1 to 5 percent band look best because it also had the most quotations, the win rate column makes the deepest band look best, and only the ratio of the two tells the truth.

Figure 9.1 draws both columns together, because the shape is easier to hold in mind than the numbers: four indigo bars that rise steadily, and an orange line above them that rises once and then bends down, with its peak marked at the second band.

Figure 9.1 — One year of quotations by discount band: the win rate rises while the margin per quotation sent peaks. The horizontal axis carries the four discount bands of Table 9.4, from no discount to more than ten percent. The indigo bars, read on the left axis, give the win rate of each band, and they rise without interruption from 22.58 percent to 39.29 percent, 41.67 percent and 46.15 percent. The orange line with square markers, read on the right axis in dollars, gives the margin earned for every quotation sent into the band, and it rises from $609.68 to a peak of $1,004.14 in the 1 to 5 percent band before falling to $965.00 and $936.00. The peak is marked, because it is the point at which a deeper discount begins to cost more margin than it buys in wins. Under each band the figure prints the quotations sent, the quotations won and the unit margin left after the discount. The panels below record the best band and the year's realization: $148,848.00 kept of $163,200.00 listed, after $4,896.00 granted off the invoice. Hydrosentinelle inc. is fictional, and every figure is computed by one model.

9.5.3 The arithmetic that finds the band where discounting stops paying

The finding above is not an accident of these particular numbers, and the arithmetic behind it generalizes. A deeper discount lowers the margin on every deal it wins, so it has to raise the win rate by a proportional amount merely to stay level. The required increase is the ratio of the two unit margins, which is the break-even volume change of Chapter 2 applied to a population of quotations instead of to a single price move.

The table that follows performs that test on the case. Its five columns give the band, the unit margin left after its discount, the win rate the band would have needed in order to match the best band's $1,004.14 per quotation sent, the win rate it actually achieved, and the verdict. The comparison is deliberately made against the best band rather than against zero, because the question a business faces is never whether to discount but how deep to go.

Table 9.5 — The win rate each band would have needed to match the best band

Three readings follow from that table, and the third is the one to act on.

The first is that the no-discount band is the worst performer, by a wide margin. It would have needed a win rate of 37.19 percent and achieved 22.58 percent, so a policy of never discounting would have cost this business more than any discount it granted. Refusing to discount is a position, not a policy, and on these numbers it is an expensive one.

The second is that the first step into discounting paid handsomely. Moving from no discount to the 1 to 5 percent band costs 5.33 percent of the unit margin, which required the win rate to rise from 22.58 percent to only 23.85 percent to break even; it actually rose to 39.29 percent. That step bought far more than it cost.

The third is that every step after the first failed the same test. Moving from the 1 to 5 percent band to the 6 to 10 percent band required a win rate of 43.36 percent and produced 41.67 percent; moving from there to the deepest band required 47.59 percent and produced 46.15 percent. The discounts kept working in the sense that they kept winning deals, and they stopped paying, which are different statements and are confused constantly. If every band had earned at the rate of the best one, the year's margin would have been 96 times $1,004.14, which is $96,397.44 instead of $82,344.00 — $14,053.44 more, or 17.07 percent. That is an upper bound and not a forecast, since the quotations in the deepest band were not the same opportunities as those in the shallowest, and it is quoted here as the size of the prize rather than as a target.

CASE IN POINT The segment question the quotation log could not answer Fictional example; names and figures are invented. When Karim Haddad brought the band reading to the president, Laurence Bergeron asked the obvious next question: who was in each band? The intuition in the room was that the deep discounts had gone to the large portfolio managers, who negotiate hard and buy many buildings at once, and that the seven full-price wins had come from single-building owners who compare nothing. That intuition may well be right, and the log could not confirm it, because the spreadsheet had a column for the discount and no column for the kind of buyer. The company could see that 13 quotations had carried more than ten percent and that 6 of them were won; it could not see whether those thirteen were thirteen large portfolios or three large portfolios and ten owners who simply asked. The answer was not to guess. Karim Haddad added one column to the log, with four permitted values — single-building owner, small portfolio manager, large portfolio manager, institution — filled in when the quotation is prepared rather than afterwards, and a second column for the competitor named on any loss. Nothing else changed. The cost was under an hour, the benefit arrives in twelve months, and in the meantime the company holds a measured finding about bands and an explicit blank where the segment finding will go. That blank is worth more than an assumption dressed as a fact, and a reader who recognizes their own log in this description should do the same thing this week rather than after the next disappointing quarter.

9.5.4 What to do with the reading on Monday

A distribution read once and admired changes nothing. Four actions follow from a reading of this kind, and they belong to chapters this book has already written.

• Move the authority ladder to where the money is. If the evidence says that reductions beyond a certain band stop paying, that band is where approval should become genuinely difficult, and Chapter 6 supplies the mechanism.
• Convert a band everybody reaches into a published structure. A discount that two-thirds of buyers obtain is not a discount; it is the price, and it belongs on the price list as a volume band or a term band with a fence around it, which is Chapter 5's work.
• Attack the empty band rather than the deep one. On the case, the striking weakness is not the deepest band but the no-discount band, whose win rate of 22.58 percent suggests that those quotations were being sent into situations where the offer was not competitive or the buyer was not qualified, which is a matter for Book E1 and Book E2 rather than for the price list.
• Repeat the reading before drawing a trend. One year of 96 quotations is a small sample, and the difference between 41.67 percent and 46.15 percent is a handful of deals; the reading is strong enough to act on because it is consistent in shape across four bands, and it is not strong enough to publish as a law.

9.6 Measuring Willingness to Pay Without Asking the Question

9.6.1 Why the direct question fails

Book A1 (From Idea to Business Model), Chapter 7, promised in this book's name that A7 would supply ways of testing a price hypothesis that do not depend on asking people what they would pay. This section is that promise. It begins with why the question fails, because a reader who does not believe it will keep asking it.

The naive question — what would you pay for this? — produces answers that are wrong in both directions and wrong for different reasons. A respondent who likes the person asking, or who wants to encourage a new venture, overstates. A respondent who suspects the answer will become the price understates. A respondent who has never faced the decision has no answer at all and produces one anyway, because the question demands a number and social convention supplies one. None of these errors cancels the others, and all of them arrive with a confidence that makes the answers difficult to discard once they are written in a slide.

The deeper problem is that willingness to pay is not an opinion people hold and can report. Chapter 2 defined it as the highest amount a buyer would actually hand over rather than take the alternative, and showed that it sits below the economic value of the offer for four reasons a survey cannot capture: doubt about whether the promised effect will occur, switching and disruption costs the buyer counts and the seller forgets, the budget authority of the person answering, and inertia. Those four live in behavior, not in memory, which is why every method below observes a behavior or constructs a trade-off rather than requesting a valuation.

9.6.2 Four instruments, and what each is worth

Measuring price sensitivity is not a fringe activity within the discipline, whatever its absence from small-business practice may suggest: of the twelve chapters of the seventh edition of Nagle et al. (2023), one is devoted to the measurement of price sensitivity and another to financial analysis, which places measurement beside price level and price structure rather than after them. The table that follows sets out the four instruments available to a business of this size. Its four columns give the method, what it actually observes, what it costs to run, and what it cannot tell you. The last column is the one to read first: no instrument here answers every question, and choosing one without knowing its blind spot is how a business ends up with a confident wrong price.

Table 9.6 — Four ways to measure willingness to pay without asking the question

The first two cost nothing and are skipped almost universally. Reading past purchases means asking what the buyer already spends on the problem: a property manager who pays for inspections, carries a deductible and has settled two water claims has revealed a budget, and that revealed budget is stronger evidence than any stated intention. Reading the discounts already granted is Section 9.5, and it deserves the name of a measurement method because that is what it is — a natural experiment that has already run at the company's expense, whose results are sitting in a folder. The caution that travels with it is important and is stated in Table 9.6: the bands were not randomly assigned. The deals that received deep discounts were probably the harder deals, so part of the difference between bands reflects the deals rather than the discounts. That is a reason to read the result as strong evidence rather than proof, and not a reason to ignore the only experiment the business has ever run.

Conjoint analysis shows buyers complete offers that differ on several attributes, price among them, and asks them to choose between offers rather than to value one. Because the respondent must give something up to get something, the method recovers what each attribute is worth in money and is much harder to game than a direct question. Green and Srinivasan (1990) review the methodological developments of the technique and treat the practical problem that limits it in a small business, which is how to estimate preferences when the number of attributes exceeds what any respondent can evaluate; they also discuss the choice simulators through which a study is converted into predicted preference shares for product and price configurations that were never shown as such. Two things follow for a reader of this book. Conjoint analysis is a genuine instrument and not a survey in disguise. And it is an instrument a small business buys once, for a decision large enough to justify the expense — a new tier structure, a new product line, a move into a new segment — rather than a routine part of pricing.

9.6.3 The price sensitivity meter, and the limits that come with it

The fourth instrument is the price sensitivity meter, introduced by Van Westendorp (1976) at the ESOMAR congress of that year as a technique for the measurement of price perception in surveys. It asks a respondent four questions about the same offer and reads an acceptable range from the points where the resulting curves cross.

A precision about sourcing belongs here, and this book states it rather than hiding it. The 1976 paper is available only to members of the association that published it and was not read for this book; only the catalog notice was. The four questions below are therefore written in this book's own words, as a faithful description of a method that has been in general use for fifty years, and they are not presented as a quotation from Van Westendorp. A reader who intends to publish results from the method should obtain the original and use its wording.

• The first question asks at what price the offer would seem so low that the buyer would doubt its quality.
• The second asks at what price the offer would seem a bargain, cheap but not suspiciously so.
• The third asks at what price the offer would start to seem expensive, though still worth considering.
• The fourth asks at what price the offer would be too expensive to consider at all.

The method's attraction is that no question asks for a valuation; each asks for a boundary, which respondents find easier and answer more consistently. Its limits are severe and must travel with any result. It measures perception rather than behavior, and a respondent who says a price is acceptable has not paid it. It is sensitive to the wording and to the sample, so two studies of the same offer can produce different ranges. It presumes the respondent understands the offer, which is rarely true of anything new. And it produces a range rather than a price, which is exactly what it is good for: bounding the region a field test should probe. Used to bound a range it is worth the afternoon; used to choose a price it is worth nothing.

WHAT CHANGED Measured price sensitivity has grown over four decades, and an old estimate understates it A business tempted to look up an elasticity for its category should know what has happened to those measurements over time. A meta-analysis of 1,851 price elasticities drawn from 81 studies reports an average of −2.62, meaning that a one percent increase in price went with a fall in quantity of about 2.6 percent, and reports three findings about the corpus itself: sales elasticities have grown in absolute size over four decades, while market-share and choice elasticities have remained stable; accounting for the endogeneity of price substantially increases the magnitudes, so studies that ignore it understate sensitivity; and accounting for consumer heterogeneity changes very little (Bijmolt et al., 2005). The practical consequence is a rule about borrowed numbers rather than a number to borrow. An elasticity taken from an older study, or from a study that did not correct for endogeneity, understates how sensitive buyers are today, so a business planning a price increase on such a figure is planning on the optimistic end of the evidence. The −2.62 itself must not be transposed to the case or to any reader's business: it is an average of a body of work dominated by consumer packaged goods measured with scanner data, and it is not an elasticity for a service, for software or for a small business. It is a reminder that demand moves, and an argument for measuring your own response rather than importing someone else's.

9.6.4 Revising the price hypothesis

Book A1 (From Idea to Business Model), Chapter 7, forms one defensible price hypothesis per segment and hands the rest of the work to this book. The instruments above are how that hypothesis is revised, and revision has a shape that is worth naming, because a hypothesis revised continuously is not a hypothesis and a hypothesis never revised is a belief.

A price hypothesis is revised when evidence accumulates on one side of it, and the evidence comes in a fixed order of strength. The weakest evidence is what buyers say. Stronger is what buyers choose in a constructed trade-off, which is conjoint analysis. Stronger again is what buyers did at prices actually charged, which is the band reading. Strongest is what buyers did at a price deliberately varied, which is a field test. A revision justified only by the weakest level should not be made; a revision supported at the third or fourth level should be made promptly, because the evidence took a year to accumulate and will not improve by waiting.

The revision itself is bounded. It changes one thing — the level of one tier, the depth of one band, the minimum, the term — and it is written down with the evidence that produced it, the date it takes effect and the measurement that will judge it. Chapter 8 governs how a change reaches existing customers, with its notice period, its grandfathering and its tolerable loss; Chapter 10 puts the revision on a calendar. What this section adds is the discipline of writing the expected outcome before the change, because a business that does not record what it expected will find, a year later, that it expected whatever happened.

9.7 Price Testing

9.7.1 What a test has to have before it is a test

A price test is an experiment in which one group is charged one price and a comparable group is charged another, and the outcomes of the two groups are compared. Four conditions make it an experiment rather than an anecdote, and a business that cannot meet all four should not describe what it is doing as a test.

• Two groups, not one period after another. The most common failure is to change the price in March and compare April with February, which measures the season, the pipeline, the weather and the price, in unknown proportions.
• Comparable groups. If the higher price goes to the buyers the business likes best, the test measures the buyers. Assignment must be by something unrelated to the outcome: a postal region, a calendar week, an arbitrary customer identifier, a channel that serves comparable buyers.
• A pre-declared outcome and duration. What will be measured — win rate, units, revenue, margin per quotation sent — and for how long, written before the first quotation goes out. A test that runs until the result looks good has no result.
• A sample large enough for the effect being sought. This is the condition that ends most small-business price tests before they begin, and Section 9.7.2 computes it.

WATCH OUT A test with no control group measures the season The before-and-after comparison is the single most common way a small business convinces itself of something false about its prices. It fails because the period before a change and the period after it differ in every way a business can differ: the pipeline that happened to be open, the season, a competitor's campaign, a large customer's budget cycle, the weather in a trade that installs equipment in buildings, and the confidence of the person quoting, which is never higher than in the weeks after a decision they argued for. All of those move a win rate by more than a price change does, and none of them is visible in the comparison. The failure is worse when the result confirms the hope, because a confirmed hope is never re-examined. If a genuine control group is impossible — and for many businesses it is — then say what the comparison is: a before-and-after observation, weak evidence, consistent or inconsistent with the hypothesis, to be revisited after another period. That sentence costs nothing and keeps a business from building a price list on a season.

9.7.2 The sample an honest test needs, computed on the case

Here is the number this book exists to give a reader, and it will not be softened. Suppose Hydrosentinelle wanted to know whether a change in price moves its win rate by 5 percentage points — from 35.42 percent to about 40.42 percent — which is a large effect, worth roughly a tenth of the year's won deals. Detecting it with the ordinary conventions of statistical work, meaning 95 percent confidence and 80 percent power, would require 1,477 quotations in each arm of the test, 2,954 in total. At 96 quotations a year, that is 30.8 years of quoting.

The company will not be running that test. Neither will the reader.

The table that follows generalizes the result, because the honest question is not whether a test is possible but how large an effect a business would have to be hunting for a test to be feasible. Its four columns give the effect to be detected, the quotations needed in each arm, the total, and the years of quoting that total represents at the case's volume of 96 a year. All four rows use the same conventions of confidence and power as the calculation above.

Table 9.7 — The quotations a two-arm price test would need, at the case's volume

Read the table from the bottom. A business at this volume can detect an enormous effect — a change that moves the win rate by a quarter of its range — in about sixteen months. It can detect a large effect in two years. It cannot detect a moderate effect in any horizon a small business plans over. Turned around, the same arithmetic says that splitting a single year of 96 quotations into two arms of 48 gives the power to detect a difference of about 28 percentage points and nothing smaller, which is to say that such a test can only discover a catastrophe.

The conclusion is the one this chapter refuses to dress up: a small business at this volume cannot test a price by running an A/B test, and should stop planning to. That sentence is not a counsel of despair, and it is not an argument against evidence. It is an argument about which evidence is available. What remains available, and is used throughout this chapter, is the analysis of margin per quotation sent across the bands the business has already granted, a sequential test on a single segment where the expected effect is large, the structured analysis of losses, and the judgment of people who read those three things regularly. That is a smaller toolkit than a large retailer's, and it is the true one.

9.7.3 What works instead, at ninety-six quotations a year

Four practices do work at this volume, and each of them accepts weaker evidence in exchange for being executable.

• The sequential test on one segment. Change the price for one segment, one channel or new customers only, choose an effect large enough to see, and run it until a pre-declared number of quotations has gone out rather than until the result pleases. The evidence is weaker than a randomized test and it is far better than a memory.
• The natural experiment already in the log. The discount bands of Section 9.5 are prices that were actually charged to real buyers, and reading them costs nothing. Their weakness is that assignment was not random, and naming that weakness is part of using them.
• The bounded increase with a measured loss. Chapter 8 raised the subscription from $79.00 to $89.00 with sixty days of notice and six months of grandfathering, computed a tolerable loss of 28 subscriptions, and lost 6. That is an experiment with a pre-declared threshold, and it is the most practical form of price evidence a small business ever generates.
• The qualitative loss review. Ten losses read carefully, with the competing price and the winner named, tell a business more than a hundred rows of a field it does not trust, and they are the input to the price-or-targeting diagnosis of Section 9.3.

What a good field experiment actually looks like is worth seeing once, because it calibrates expectations. Anderson and Simester (2003) tested nine-ending prices in two women's clothing mail-order catalogs by randomizing versions across postal codes and measuring units sold, with samples running to tens of thousands of customers per version and hundreds of item-level observations per study, deducting returns and cancellations; the effects they measured ranged from about 7 percent to about 35 percent depending on catalog and context, and the authors described one of their own comparisons as a tentative finding worth further investigation. That is the scale and the modesty of a real price experiment. A business sending eight quotations a month is not in that world, and the useful response is to choose different instruments rather than to imitate the vocabulary.

METHOD Running a price test that answers a question When to use it: when a specific price decision is pending, the expected effect is large, and one group of buyers can be charged differently from another without unfairness to either — typically a new segment, a new channel, or new customers only, never two identical buyers in the same market on the same day. Steps: (1) Write the question as a single sentence naming one price, one group and one outcome, such as whether the entry tier at a higher monthly price wins the same share of single-building owners. (2) Compute the sample the effect requires, using the arithmetic of Table 9.7, and abandon the test now if the sample exceeds what the business will generate within the decision's horizon. (3) Choose the assignment rule before anything else, so that group membership is decided by something unrelated to how promising the buyer is, and write down who is excluded. (4) Fix the duration or the number of quotations in advance, and record both in the same document as the question. (5) Declare the outcome measure and the threshold that will change the decision, preferring margin per quotation sent over win rate, because win rate can be bought. (6) Keep every other variable still: no promotion, no new collateral, no change to the quotation template while the test runs. (7) Record each quotation in the log with its arm, its price and its outcome, including the quotations that expire. (8) At the end, compute the outcome for both arms, state the difference in plain numbers, and state explicitly whether the sample was large enough for the difference to mean anything. Deliverable: a one-page test record carrying the question, the assignment rule, the sample computed in advance, the dates, the two results and a written conclusion that names the decision taken. Verification criterion: someone who was not involved can read the page, recompute the required sample, see that the actual sample met or missed it, and reach the same conclusion from the same numbers.

THE RULE Competition Act, s. 74.1(3): a due-diligence defense is built before anyone asks for it Section 11.6 reads the same section for who is liable; this one reads it for what the records must show. Part VII.1 of the Competition Act, RSC 1985, c C-34 (Canada) treats a false or misleading representation to the public as reviewable conduct, without any need to prove intent. Section 74.1(1)(c) sets the administrative monetary penalties a court may order: for a corporation, on a first order, the greater of $10,000,000 and three times the benefit derived from the conduct, or, where that benefit cannot be reasonably determined, 3 percent of annual worldwide gross revenues, rising on a subsequent order to the greater of $15,000,000 and three times the benefit; for an individual, the greater of $750,000 and three times the benefit on a first order. Section 74.1(1)(d) permits an order of restitution to purchasers. Two further provisions matter to a chapter about measurement. Section 74.1(3) provides that no penalty is to be imposed on a person who establishes that they exercised due diligence to prevent the conduct, and section 74.1(5) lists the factors a court weighs, among them the materiality of the representation, the effect on competition and the person's history of compliance. Neither provision is satisfied by good intentions. What satisfies them is contemporaneous evidence, and the records this chapter asks a business to keep — the dated price list in force, the quotation log, the evidence of what was actually charged and when — are the raw material from which such a defense is assembled. None of them can be reconstructed after a question has been asked. Testing a price changes nothing about this: a price charged during a test is a real price, and a representation made about it is a real representation, reviewable under section 74.01 and, where it is made knowingly or recklessly, an offense under section 52. Jurisdiction: Canada, federal, applying to conduct in Canada. Justice Laws Website, current to July 21, 2026, last amended March 26, 2026, read September 19, 2026. Verify the current text before relying on it.

9.7.4 Dynamic pricing, and the data it demands

Book A1 (From Idea to Business Model), Chapter 7, separates a fixed price from a dynamic price — one that varies automatically with demand, time, inventory, channel or the identity of the buyer — and declares dynamic pricing out of reach for a business at the discovery stage. This book agrees, and owes the reader the reason, because the phrase is used loosely enough that owners think they are missing something.

A dynamic price is not a price that changes; every price in this book changes, and Chapter 8 governs how. A dynamic price is a rule that changes the price without a human decision, and a rule of that kind requires five things the business must have before it can be written.

• Volume, because a rule is fitted to observed responses and a business with 96 transactions a year cannot fit anything; the sample arithmetic of Table 9.7 is the same obstacle in a different costume.
• A measured demand response, meaning the business knows how quantity moves with price in its own market, which Chapter 2 explains a small business cannot look up and this chapter explains it usually cannot measure either.
• A perishable or capacity-constrained inventory, since the economic logic of dynamic pricing is that an unsold seat, night or hour is worth nothing tomorrow; equipment and monitoring subscriptions are not perishable in that sense.
• A transaction system that can apply and record the rule, including which price applied to which buyer at which moment, because without that record neither the business nor a regulator can reconstruct what was charged.
• A published basis that the buyer can understand, because a price that changes for reasons the buyer cannot see is the fastest way to destroy the reference price Chapter 3 describes, and because everything said in Chapters 3, 6 and 7 about representations applies to each price the rule produces.

Three things that are often called dynamic pricing are not, and they are available to any business. Seasonal or time-of-week pricing is a published schedule, decided by a person, which a buyer can read in advance. Volume and term bands are published structures, and Chapter 5 built them. Segment pricing is a published fence, also Chapter 5's work. Each of these captures part of what dynamic pricing promises, is legible to the buyer, and requires no data the business does not already have. The honest statement for a reader of this book is that dynamic pricing is a legitimate technique in industries with perishable capacity and thousands of daily transactions, and that in a small business the phrase almost always describes either a published schedule or an undisclosed practice of charging different buyers different prices with no fence, which Chapter 5 and Chapter 6 both warn against.

9.8 The Quarterly Pricing Review

9.8.1 The agenda

Measurement that is not attached to a recurring meeting stops after two quarters. The pricing review is that meeting: ninety minutes, four times a year, with a fixed agenda, a named owner and a written output. Chapter 10 places it in the calendar of the pricing year alongside the annual review and the promotion calendar; this section says what it contains.

The table that follows is the agenda. Its four columns give each item, the measurement it reads, the decision the item is allowed to produce, and the evidence that is filed afterwards. Two features of the table matter more than its contents. Every item reads a number before it discusses an opinion, and every item has a bounded set of decisions, so that the meeting cannot drift into a general conversation about whether prices should be higher.

Table 9.8 — The quarterly pricing review agenda

9.8.2 What the review may decide, and what it may not

A review that can decide anything will decide nothing, so the boundaries are part of the design. The quarterly review may adjust the policy — authority thresholds, the floor, the discount bands, the template — because those are governance instruments and the evidence for changing them accumulates quarterly. It may exclude a segment, since the evidence for that comes from losses and arrives continuously. It may start or stop an experiment, against the threshold the experiment declared.

It may not rewrite the price list. Price levels move on the annual cadence that Chapter 10 owns, for a reason that is commercial rather than bureaucratic: a market that sees prices move four times a year learns to wait, sellers stop quoting with conviction, and every published document in circulation becomes suspect. The exception is a cost shock large enough to breach the floor, which is not a review decision but an emergency, handled with the notice and grandfathering rules of Chapter 8.

Two further constraints keep the review honest. Every decision names the measurement that will judge it and the quarter in which that judgment will happen, so that the next review opens by reading the last one. And every decision is recorded with the evidence that produced it, in a file that survives the departure of the people in the room, because the most common cause of a business relearning a pricing lesson is that the person who learned it the first time left.

QUÉBEC & CANADA NOTE What a Canadian small business can compare its own price movement against A business that measures itself asks, sooner or later, whether its behavior is normal, and for Canadian businesses one public benchmark exists. In its analysis of small businesses for the second quarter of 2026, Statistics Canada reports that 24.3 percent of businesses with 1 to 19 employees expected to increase the selling price of the goods and services they offer over the following three months, while 65.2 percent expected no change and 7.1 percent expected a decrease; among businesses with 20 to 99 employees the share expecting an increase was 31.2 percent, and among those with 100 or more employees it was 27.4 percent. The same analysis reports that 64.0 percent of the smallest group expected at least one cost-related obstacle over the same period (Statistics Canada, 2026a). Three cautions make that comparison useful rather than decorative. It measures expectations of a movement, not realized prices or realized margins, so it cannot tell a business whether its own realization is good. It is a national survey covering all industries, so a business in a single trade in a single region should treat it as background rather than as a peer group. And it is a probabilistic sample of the business population, which makes it a different kind of evidence from a business's own quotation log: the log is a census of one company and carries no sampling error at all, while the survey generalizes and cannot describe anyone in particular. Read together, they answer two different questions — what businesses in Canada expect to do, and what this business actually did. The second is the one that changes a price.

9.8.3 The first year, for a business starting from nothing

A reader who has none of this should not attempt the whole apparatus at once. The order below is the cheapest path to a real dashboard, and it has been chosen so that each step produces a usable result on its own.

The first quarter creates the quotation log and nothing else: one row per quotation sent, with the date, the buyer, the list value, the discount, the option accepted and the outcome. The second quarter adds the reconstruction of the previous year's realization from invoices and order files, which produces two numbers, invoice realization and pocket realization, and one conversation that usually changes behavior on its own. The third quarter adds the band reading of Section 9.5, which becomes possible as soon as three quarters of quotations exist and which is the reading most likely to change a policy. The fourth quarter adds the segment and loss fields, whose value arrives a year later and which therefore have to be started before anyone feels the need for them.

At the end of that year the business has six numbers, a year of history and the ability to compare itself with itself, which is the only comparison that has ever improved a price. It will also have discovered the thing every business discovers at this point: the measurement was never the hard part, and the hard part is acting on a reading that contradicts what the most confident person in the company believed.

Costly Mistakes

• Measuring realization on invoices alone. The invoice records the honest half of the concessions; the free goods, the unbilled hours and the extended terms are the half that never appears, and on the case they are the difference between 94.21 percent and 91.21 percent.
• Reporting a win rate on a denominator that excludes expiries and verbal quotations. A denominator maintained by the person whose performance it measures will shrink quietly, and every conclusion drawn from it afterwards is wrong in the same direction.
• Reading the average discount instead of the distribution. An average of six percent is produced both by a business that grants six percent to everyone and by one that grants nothing to most and eighteen percent to a few, and the two need opposite remedies.
• Concluding from a rising win rate that deeper discounts work. On the case the win rate rises across all four bands while margin per quotation sent peaks in the second, and a business that reads only the first column will keep buying wins at a loss.
• Calling a before-and-after comparison a test. Without a control group the comparison measures the season, the pipeline and the enthusiasm of the person quoting, and a result that confirms the hope is never re-examined.
• Planning an A/B test that the volume can never support. Detecting a five-point effect here needs 2,954 quotations, or 30.8 years; the time spent designing that test is time not spent reading the bands, which cost nothing and produced a finding worth $14,053.44 at the upper bound.

Key Takeaways

• Six numbers are enough: pocket realization, win rate, discount distribution, contribution by offer, mix and attach rate, and retention — each computable from records a business already has or can start keeping in an afternoon.
• Realization is the share of list value you keep, and on the case it was 94.21 percent at the invoice and 91.21 percent in the pocket, which cost 15.63 percent of the list margin against 8.79 percent of the price.
• A win rate is only as good as its denominator, and a very high one usually means the price is below the point where buyers hesitate rather than that the selling is excellent.
• A loss to a structurally cheaper model is a targeting signal, not a pricing signal, and the four readings of Table 9.3 tell the two apart before any price is changed.
• Discounting buys wins and stops paying for them: the case's win rate rises from 22.58 percent to 46.15 percent across the bands while margin per quotation sent peaks at $1,004.14 in the 1 to 5 percent band and falls thereafter.
• Refusing to discount is not free: the no-discount band would have needed a win rate of 37.19 percent to match the best band and achieved 22.58 percent.
• Willingness to pay is measured by observing behavior, through past purchases, the bands already granted, field tests and conjoint analysis, with the price sensitivity meter useful only for bounding a range.
• A business of this volume cannot run an A/B price test, since a five-point effect needs 1,477 quotations per arm and 30.8 years, and the honest alternatives are the band reading, the bounded increase and the sequential test on one segment.

Your Action Plan

1. Start the quotation log, or complete the one you have. The deliverable is a single sheet with one row per quotation sent, carrying the date, the buyer, the list value, the discount applied, the option recommended, the option accepted, the outcome and the loss reason; it takes thirty minutes to build and about a minute per quotation thereafter, and it is done when a full month has been recorded with no quotation missing, including those that expired.
2. Reconstruct last year's realization. The deliverable is the three totals of Table 9.2 — list value won, invoice value and pocket value — with the off-invoice estimate stated as an estimate and its basis written beside it; it takes a day with the invoices and the order files, and it is done when the two percentages are computed and compared with what your discount policy intended.
3. Compute your win rate on a definition you have written down. The deliverable is the definition in one sentence plus the rate for the last four quarters on that definition; it takes two hours, and it is done when a second person applying the definition to the same records reproduces the same number.
4. Read your year by discount band. The deliverable is Table 9.4 filled in with your own bands, aligned with your authority thresholds, including the margin per quotation sent for each band; it takes half a day, and it is done when you can name the band where margin per quotation sent peaks and say how far the next band fell short.
5. Add the two missing fields. The deliverable is a segment tag with no more than five permitted values and a loss field naming the competitor and the approximate gap, both filled in at the moment the quotation is prepared or lost; it takes an hour to define, and it is done when every quotation of a full month carries both.
6. Decide whether you can test anything, before designing a test. The deliverable is one line of arithmetic giving the sample your volume can supply in a year and the smallest effect that sample could detect, using Table 9.7 as the model; it takes fifteen minutes, and it is done when you have written either a test design with a sample that fits or a sentence saying that a test is not available to you at this volume.
7. Put the review in the calendar four times. The deliverable is four ninety-minute meetings booked for the next year with the agenda of Table 9.8 attached, an owner named and a file opened for the written outputs; it takes fifteen minutes, and it is done when the first one has taken place and produced a decision with the measurement that will judge it named beside it.

In La Machine

Every route named here was verified in the route inventory on September 19, 2026, and the behaviors were read in the application's source at its commit of September 7, 2026 (La Machine, 2026); Book N8 will cover these pages one by one, and this section says only what a reader trying to measure pricing performance can and cannot obtain from the platform today.

• Revenues & Expenses (business/subscriptions, reached as Account → Profile → Revenues & Expenses) reads the subscription transactions, the package purchases and the product purchases and writes each sale as a revenue line with its cost lines, which makes it the only view in the product that places a revenue beside what it cost; the sample data are generated, with sales taxes of 5 percent and 9.975 percent written into the generator, so a reader should confirm which rows are real before drawing anything from them.
• Business → Session Packages (business/packages, also reached as productivity/packages) and the subscription packages page (business/subscription-packages) each hold a price and a cost and display profit per sale and a margin percentage on screen without storing them, which answers the contribution-by-offer line of Table 9.1 one offer at a time.
• Business → Store → Products (business/shop, business/shop-product-edit) holds the product cost, the shipping cost and the shipping fees with a profit panel, so a product's contribution is visible on the page where its price is set.
• Control center gives the cadence rather than the numbers: Projects (chains) can carry the quarterly review as a recurring project, Plan tasks and Today can hold its preparation, and Audit log records who changed what, which is the trail the review needs and Chapter 10 develops.
• What is not here. There is no quotation object at all, since business/quotes is a page of motivational quotes, so there is no quotation log, no denominator and therefore no win rate; there is no price realization, no pocket price, no off-invoice concession field and no price waterfall; there is no discount object beyond a whole-number percentage on a promotion with a start and an end date, so no discount distribution and no discount authority; there is no price test and no A/B capability for a price, no elasticity, no price optimization and no forecast; there is no competitor record; there is no margin floor and no price list with an effective date; the /dashboard route displays demonstration widgets and carries no revenue, margin or discount indicator; and the platform checks no price presentation against the Competition Act or the Consumer Protection Act, which Book B1 (Marketing Fundamentals for Business Owners), Chapter 6, states already and this chapter confirms. Every measurement in this chapter is therefore produced outside the platform, from the platform's own exports and from the business's invoices and order files.

Self-Check

1. What are the three values a realization calculation needs, and why is the first of them the one businesses never have?
2. On the case, which discount band earned the most for every quotation sent, and what does the win rate column say that contradicts it?
3. What win rate would the no-discount band have needed to match the best band, and what does the answer say about a policy of never discounting?
4. How many quotations would a two-arm test need to detect a five-point change in the win rate, and what should a business at this volume do instead?
5. What distinguishes a price problem from a targeting problem, and which two fields must exist in the log before the question can be answered?

Answers. (1) The list value of what was won, the invoice value and the pocket value; the list value is the one no document contains, because nothing in the accounting system records what would have been charged at the published price, and its absence is why realization goes unmeasured for years (Sections 9.2.1 and 9.2.2). (2) The 1 to 5 percent band, at $1,004.14 for every quotation sent, against $609.68, $965.00 and $936.00; the win rate column rises without exception from 22.58 percent to 46.15 percent, so discounting does keep buying deals, and the two columns together show that it stops paying for them after the second band (Sections 9.5.2 and 9.5.3). (3) It would have needed 37.19 percent and achieved 22.58 percent, short by 14.61 points, which makes the no-discount band the worst performer of the four and a policy of never discounting the most expensive position available on these numbers (Section 9.5.3). (4) 1,477 in each arm and 2,954 in total, which is 30.8 years at 96 quotations a year; instead, the business reads the bands it has already granted, runs bounded increases with a pre-declared tolerable loss, runs sequential tests on one segment where the expected effect is large, and reviews its losses qualitatively (Sections 9.7.2 and 9.7.3). (5) A price problem shows up as losses to comparable competitors at small gaps, spread across segments; a targeting problem shows up as losses to a structurally cheaper offer at gaps no discount could close, concentrated in one segment — and answering it requires a segment tag and a loss field naming the competitor and the approximate gap (Sections 9.3.3 and 9.5.3).

Chapter References

Anderson, E. T., & Simester, D. I. (2003). Effects of $9 price endings on retail sales: Evidence from field experiments. Quantitative Marketing and Economics, 1(1), 93-110. https://doi.org/10.1023/A:1023581927405 (full text read September 19, 2026, in the copy deposited by the author at the Kellogg School of Management, Northwestern University)

Bijmolt, T. H. A., Van Heerde, H. J., & Pieters, R. G. M. (2005). New empirical generalizations on the determinants of price elasticity. Journal of Marketing Research, 42(2), 141-156. https://doi.org/10.1509/jmkr.42.2.141.62296 (abstract and metadata read September 19, 2026, at the University of Groningen research portal)

ChartMogul. (2025, December 10). The SaaS retention report: The AI churn wave. Retrieved September 19, 2026, from https://chartmogul.com/reports/saas-retention-the-ai-churn-wave/ (annualized 2025 rates from approximately 3,500 vendor websites categorized with the assistance of artificial intelligence; a non-probability sample)

Competition Act, RSC 1985, c C-34, ss. 52, 74.01 and 74.1 (Canada). Justice Laws Website, current to July 21, 2026, last amended March 26, 2026. Read September 19, 2026.

Faustino, S. (2025, May 18). SaaS billing report: Billing is the invisible force behind your growth. ChartMogul. Retrieved September 19, 2026, from https://chartmogul.com/reports/saas-billing-report/ (aggregated billing data from more than 2,500 subscription software companies for full-year 2024; a non-probability sample of one vendor's own customers, excluding companies below $300,000 of annual recurring revenue)

Green, P. E., & Srinivasan, V. (1990). Conjoint analysis in marketing: New developments with implications for research and practice. Journal of Marketing, 54(4), 3-19. https://doi.org/10.1177/002224299005400402 (notice and abstract read September 19, 2026, at the Stanford Graduate School of Business publications page)

La Machine. (2026). Documentation — route reference inventory (211 routes: 191 Core, 20 Pilotage paths), and the product source code read at commit be8bf04 of September 7, 2026. Internal product documentation, read September 19, 2026.

Marn, M. V., & Rosiello, R. L. (1992, September–October). Managing price, gaining profit. Harvard Business Review. https://hbr.org/1992/09/managing-price-gaining-profit (article page and publisher product description read September 19, 2026; the full text was not read, and the pagination reported by secondary sources could not be verified at source)

Nagle, T. T., Müller, G., & Gruyaert, E. (2023). The strategy and tactics of pricing: A guide to growing more profitably (7th ed.). Routledge. https://doi.org/10.4324/9781003179566 (publisher page read September 19, 2026)

Statistics Canada. (2026a, June 18). Analysis on small businesses in Canada, second quarter of 2026 (Catalogue no. 11-621-M). Retrieved September 19, 2026, from https://www150.statcan.gc.ca/n1/pub/11-621-m/11-621-m2026008-eng.htm

Van Westendorp, P. H. (1976, September). NSS-Price Sensitivity Meter (PSM): A new approach to study consumer perception of prices [Paper presentation]. ESOMAR 29th Congress, European Society for Opinion and Marketing Research. https://ana.esomar.org/documents/nss-pricesensitivity-meter-psm- (catalogue notice read September 19, 2026; the paper itself is restricted to members of the association and was not read, so the four questions described in Section 9.6.3 are this book's own wording and not a quotation)

Chapter 10 — Typical Tasks, Cadence and Workflows

On the last Friday of the third year's first quarter, Laurence Bergeron wrote the year's pricing decisions on a single sheet and discovered that every one of them was finished and none of them was running. The three tiers existed, at $89, $129 and $179. The floor existed, at $3,500, with an authority of 6 percent for a representative and 12 percent for the head of sales. The quotation to Groupe immobilier Vallières had gone out with its taxes on their own lines, its deposit of $16,243.20 and its validity of 30 days. The subscription increase from $79 to $89 had been decided, and nobody had yet agreed on the day the notice would go out. What the sheet did not contain was a single recurring date, a single owner other than the president, and a single sentence saying how anyone would know that a pricing task had been done. Hydrosentinelle had spent a year deciding prices and had not yet built the year that keeps them. This chapter is that year. It turns the pricing work into a task table in the ten fields La Machine uses, sets what recurs every day, week, month, quarter and year, runs the annual price review that governs all of it, and writes the three workflows a small business actually needs — the quotation, the discount approval and the promotion — each with a verification criterion and a record.

The running case of this book, Hydrosentinelle inc., is fictional; its names and figures are invented and computed by a single model.

What You'll Be Able to Do

• Write your pricing work as a task table in the ten fields La Machine uses, with a trigger, a frequency, an owner, an output and a verification criterion for every line.
• Separate the standing decisions from the recurring reads and from the event-driven moves, and give each kind the cadence it actually needs.
• Build a pricing calendar for a year, from the daily discount answer to the January review, and place a price increase inside it with its notice period and its grandfather window.
• Count the hours your pricing work costs, in a steady year and in a year that carries an increase, and decide which of those hours you will delegate first.
• Run the three workflows of pricing — the quotation, the discount approval and the promotion — each with the stage at which it may be stopped and the record it must leave.
• Write a verification criterion for every pricing task that someone other than the person who did it can check, and keep the records that Chapter 9 measures from.

Why It Matters

A price is the only decision in a business that erodes silently when nothing happens. A supplier who raises a cost sends an invoice; an employee who leaves says so; a customer who is unhappy complains or disappears. A price that has not been reviewed sends nothing at all, and the erosion arrives as a slightly lower margin rate spread across a year and attributable to nothing in particular. That is why pricing needs a calendar more than almost any other function: the work is not difficult, it is simply never urgent, and everything that is never urgent is done by a date or not at all.

For a solo owner, the calendar is short and its value is that it exists. A monthly half-hour with the margin, a quarterly hour with the discounts granted, an afternoon in January with the price list, and two dated promotion windows are enough to keep a price from drifting for a decade. The solo owner's specific risk is that the person who sets the price and the person who gives it away are the same person, so the calendar is the only separation available: the floor written on a scheduled day in January is what the same person must obey in March. For a small team, as Hydrosentinelle now is, the calendar is what allows pricing work to be shared without being lost. The moment two people can quote, the business needs a same-day answer to a discount request, a weekly log of what went out, and a monthly reading of what was actually kept; without those three, the price list becomes a fiction within a quarter and nobody can say when it stopped being true. For an established business, the calendar becomes governance: a named owner for the price, a review that produces a dated decision, an approval trail a later auditor can follow, and a record of every price change with the person who made it. The strategic pricing literature treats that capability as a thing a company builds deliberately rather than a thing that appears, and it places the building of it alongside the analysis itself (Nagle et al., 2023).

There is also a legal reason for the dates, and it is the reason this chapter carries a rule of its own. Several of the obligations that bind a displayed price are obligations about time: a price advertised for a period must be supplied at that price during that period, an ordinary-price comparison must rest on what was really charged over months, and a promotion that outlives its published end date quietly turns the reduced price into the ordinary price. Those are calendar problems before they are legal problems, and a business that keeps the calendar rarely meets the law from the wrong side.

In This Chapter

• 10.1 The Pricing Calendar of a Year
• 10.2 What Happens Every Week
• 10.3 What Happens Every Month
• 10.4 What Happens Every Quarter
• 10.5 The Annual Price Review
• 10.6 The Quote Workflow
• 10.7 The Discount Approval Workflow
• 10.8 The Promotion Workflow

10.1 The Pricing Calendar of a Year

10.1.1 Three kinds of pricing work, and the way each one fails

The pricing work of a business divides into three kinds, and the division matters because each kind fails differently and each needs a different sort of date. Standing decisions are written once and revised on a schedule: the price list with its effective date, the tier structure, the margin floor, the authority ladder, the list of acceptable trades and the promotional calendar. They are the output of Chapters 2 to 8, and they are what the annual review of Section 10.5 reopens. Recurring reads come back on a fixed cadence whether or not anything has happened: the quotation reviewed before it is sent, the discount answered the same day, the week's quotations logged, the month's margin and realization read, the quarter's discount bands examined. Event-driven moves are triggered by something that occurs: a supplier cost that jumps, a promotion window that opens, an increase that is decided, a request to go below the floor.

Each kind fails in its own way. Standing decisions fail by silent expiry: the price list is still on the shared drive, it still looks authoritative, and it has been wrong since the last cost increase. Recurring reads fail by being skipped, and they are skipped in a specific order — the monthly margin read goes first, because nothing visible happens when it is missed, and the business then loses the only instrument that would have told it the price list had expired. Event-driven moves fail because nobody recognizes the event: the cost moved in February, everyone noticed in the sense of seeing the invoice, and no one had written the sentence that says a cost movement above a stated threshold triggers a re-pricing within two weeks.

DEFINITION Pricing governance, pricing calendar, price change request and the definition of done Pricing governance is the set of arrangements that decide who may set or change a price, on what evidence, by what procedure and with what record; it is a small thing in a small business and a named function in a large one, but it exists in both, either written down or improvised. A pricing calendar is the dated list of every recurring pricing task, with its trigger, its frequency, its owner, its output and its verification criterion, plus the dated events of the year such as the review, the promotion windows and the effective date of any change. A price change request is the written form in which a proposed departure from the published price travels to the person entitled to decide it: what is asked, for whom, what is obtained in exchange, what margin results, and who approved it. A cadence is simply how often a task recurs, stated as a frequency rather than as an intention. And a definition of done, which Book J2 (Processes, SOPs and the Definition of Done), Chapter 4, develops in full, is the observable condition that says a task is finished; in this book it appears as the verification criterion column of every task, and it is written so that a person who did not do the work can check it from a record rather than from a report. In plain terms: governance says who decides, the calendar says when, the request says how a departure travels, and the verification criterion says how anyone else will know the work was really done.

10.1.2 The task table, and why it fits in six printed columns

La Machine describes every task in ten fields: the task, its objective, its trigger, its frequency, its automation level, its input, its output, its verification criterion, its role and its tools. Book J2, Chapter 8, uses that format for a documentation program, and Books E1 and E8 use it for the sales process and for sales management. Ten fields do not fit across a printed page, so the table below pairs them: the objective follows the task name in italics, the trigger, the frequency and the automation level share one column, the input and the output share the next, the verification criterion stands alone because it is the column that makes the table checkable, and the role and the tools share the last. Nothing is lost in the pairing, and the table can be read aloud line by line.

The hours in the task column are the founders' own hours, estimated by this chapter rather than computed from the case model, on two stated conventions: 48 working weeks and 240 working days in a year. They are there because the cost of running a price is one of the things a small business consistently fails to budget, and because the delegation decision of Section 10.1.4 cannot be made without them. The hours of an accountant or an outside adviser are not counted and not priced, since this book prints no fees of any kind.

The following table, Table 10.1, gives Hydrosentinelle's pricing work as a task table, grouped by cadence. Its six columns give the reference number, the task with its objective and its yearly hours, the trigger with the frequency and the automation level, the input and the output, the verification criterion, and the role with the tools. The rows are grouped in three blocks: fifteen recurring tasks from the day to the year, four event-driven tasks with the number expected in a year, and four workflows that software can carry in whole or in part. Read the verification criterion column first, because it is the column that turns a list of good intentions into work someone can check.

Table 10.1 — Hydrosentinelle's pricing work as a task table (format La Machine)