Price tells a story: pricing policy for decision makers

Four minutes, one Friday

Friday, 5:40 p.m. A salesperson calls the commercial manager: the big client is about to close the order of the semester, but asks for "a little help": twelve percent. On the other end there's car noise, the weekend, tiredness. The conversation lasts four minutes and ends as it almost always does: "fine, close it."

Nobody calculated anything in those four minutes. Nor did anyone know that this one-off discount would seep into the client's price list for the entire following year, or that the salesperson would use it as a precedent with two other accounts ("they got twelve"). In a company with a 30% contribution margin, that 12% discount handed over 40% of the profit on every affected sale. To recover it, volume with that client would have to grow by two thirds.

Meanwhile, in the same company, a serious committee has been meeting for months to cut freight costs by 2%.

This asymmetry —surgical rigor for costs, benign neglect for prices— is one of the strangest and best-documented regularities in management. This article is about why it happens, what evidence exists on what it costs, and how to build what most mid-sized companies lack: a pricing policy worthy of the name.

The most powerful lever on the income statement

In 1992, two McKinsey consultants, Michael Marn and Robert Rosiello, published in Harvard Business Review a calculation that became the obligatory starting point of the entire discipline. They took the average income statement of 2,463 companies and measured what a 1% improvement in each lever does to operating profit, holding everything else constant.

McKinsey repeated the exercise a decade later on the S&P 1500 and the order held: price remained the most powerful lever, with an effect of around 8% per point. The fine numbers change with each company's margin structure; the hierarchy does not: price beats volume by a multiple, and fixed costs by another. The managerial irony is evident: the variable with the greatest effect on profit is, in the typical mid-sized company, the only one with no owner, no process, and no scheduled review.

Warren Buffett put it bluntly before the Financial Crisis Inquiry Commission in 2011: "The most important decision in evaluating a business is pricing power. If you have to hold a prayer session before raising the price 10%, you've got a terrible business."

The arithmetic no salesperson carries

The flip side of the lever is the discount trap, and it deserves to be printed in every sales team's wallet. When the price drops, the volume needed to keep the same profit does not grow proportionally: it grows by the formula d/(m−d) —discount over margin minus discount— which steepens fast:

Additional volume needed to keep the same profit after a discount, by contribution margin.

At a 30% margin, a 10% discount demands selling 50% more units just to match the previous profit. How often does that happen? Experience confirms what intuition suspects: almost never. Most discounts do not buy volume; they buy the feeling of having closed.

The same arithmetic, read backwards, is the best news on the income statement: at a 30% margin, an 8% price increase that loses 3% of volume leaves variable profit 23% higher. You can lose volume and earn more — in fact, that trade, well chosen, is the very definition of a good pricing policy.

Price speaks before the salesperson

So far, the arithmetic. But the title of this article says something else, and it is the part the spreadsheet does not capture: price is a message. Before the salesperson says a word, the number has already told a story — about the product's quality, about who it is meant for, about how much the company offering it values itself.

Economic psychology documented it decades ago: buyers use price as a signal of quality when they cannot verify it directly (the literature on price–quality inference goes back to the 1960s), and we all anchor on the first number we see, even knowing it is arbitrary — the anchoring effect Tversky and Kahneman demonstrated in 1974. The wine that costs twice as much "tastes better" even in experiments with the same wine; the professional service that is too cheap generates no relief: it generates suspicion.

You do not choose whether your price tells a story; you choose only whether you are the one who writes it.

The "intuitive" price —cost plus a customary margin, the competitor's price minus a bit, the 2019 list adjusted in fits and starts— also narrates. It says the company does not quite know how much what it does is worth. And that story is told to its customers every day by its prices, with more reach than any campaign. Three typical stories told by badly managed prices:

The archaeological price list. It was born from a costing done years ago, survived three commercial managers, and gets adjusted "when there's no other choice." Each geological layer has its lost logic. It tells the story of a company that looks at its costs, not at its customer's value.

The timid price. The product improved, the service expanded, customers renew — and the price stays where it was, "so as not to make waves." It tells the story of a company worth more than it charges, and the only one who knows it is the customer.

The democratic discount. Everyone gets something if they push: the list price is merely the opening of a negotiation with no rules. It tells the story that the published price is a lie — and trains customers, one by one, never to believe it again.

Where it leaks: from list price to pocket price

Marn and Rosiello contributed the discipline's second fundamental concept: the price that matters is neither the list price nor the invoice price, but the pocket price: what actually remains after subtracting everything given up in the transaction. The tool for seeing it is the price waterfall: you start from the list price and subtract, step by step, every concession.

Two findings recur in almost every pricing diagnosis, and both tend to surprise management.

Total leakage is far greater than perceived. Each concession was approved separately, looks reasonable separately, and lives in a different system (the rebate in sales, the payment terms in finance, the freight in logistics). Nobody adds them up. In companies with low double-digit margins, it is common to discover that a quarter of the price disappears between list and pocket — and that nobody was responsible for looking at the whole thing. In inflationary contexts like ours, moreover, the payment-terms step becomes enormous: days of credit are price, and whoever does not count them is granting discounts they cannot see.

The band is enormously wide. The second finding: for the same product, in the same month, the pocket price between the best and worst customer usually differs by percentages that are shocking — and that difference is almost never explained by volume. It is explained by who negotiated, when, and how much hunger to close there was that day. It is the unmistakable signature of price without policy: the company's profitability ends up determined by the chance of individual negotiations.

Large-scale evidence confirms this is no anecdote: in Simon-Kucher's global pricing study — the discipline's largest survey, of thousands of executives across dozens of countries — companies report realizing on average less than half of the price increases they plan; in several editions, barely around a third. Between the pricing decision and the pocket price sits an entire machinery — rules, exceptions, sales incentives, silences — that decides how much survives. Most companies do not have that machinery: they have habits.

The architecture: five components

Moving from intuitive pricing to a price architecture requires neither a new department nor expensive software. It requires building five components, in this order.

1. See the real price. Everything starts with a piece of data most companies do not have: the pocket price per transaction, per customer and per product. It can usually be reconstructed from what is already in the billing system plus two or three spreadsheets. The first deliverable of any serious pricing work is that map: the average waterfall and the full band. It is the commercial equivalent of medical imaging: uncomfortable, revealing, indispensable.

2. Segment by value, not by size. Not all customers value the same things, nor do they cost the same to serve. The customer who buys urgency, technical service and availability can — and usually is willing to — pay differently from the one who buys price and says so. Willingness to pay can be measured with formal methods (Van Westendorp, conjoint analysis) or, to begin with, with the cheapest method: reading your own transactions — who pays list without arguing, who fights over every point, what was truly lost to price and what was lost to something else disguised as price. That is enough to stop charging the same to those who receive different value.

3. Rules for giving ground. The discount policy is the constitution of the architecture, and it fits on one page: what discount levels exist, who authorizes each level, and — the golden clause — every concession has a counterpart: a discount is exchanged for committed volume, for shorter terms, for mix, for an annual contract; never for insistence. The salesperson stops being a lone negotiator at 5:40 on a Friday and starts operating within a system with floor prices, tiers and fine print thought out in advance.

4. An owner and a cadence. Price needs what costs, quality and safety already have: a responsible person with a name and a routine. Someone who owns the list, the exceptions and the review calendar — monthly or quarterly depending on the industry; with inflation, the calendar is even more critical, because an outdated list is a silent, across-the-board discount that nobody approved. Every review documented: what was changed, why, what happened.

5. Measure realization. Close the loop with two or three indicators watched in the commercial routine: realized price against target price (how much of the decided increase reached the pocket?), the share of transactions outside the band, and total waterfall leakage. What is not measured erodes — and price erodes faster than almost anything, because there is an entire sales force with incentives to erode it.

On that last point, a structural warning: if salespeople earn commission on revenue, the company is paying them to give away margin. Any serious architecture ends up revising the incentive scheme so that salesperson and company win on the same variable: contribution, not volume.

The difficult conversation

One final objection remains, the one that props up the status quo: "if we touch prices, we lose customers." Three answers, in order of importance.

First: have it with the arithmetic on the table. We already saw that at a 30% margin you can lose 3% of volume after an 8% increase and end up substantially better off. The right question is never "will we lose customers?" — some churn is expected — but "how many would we have to lose for it not to be worth it?". That number is almost always surprisingly high. And the customers who leave first tend, with notable regularity, to be those at the bottom of the band: the ones who were already the least profitable in the house.

Second: the increase is communicated with value, not with costs. The letter that says "due to the increase in our costs" is an apology; the conversation that reviews what the customer received — service, availability, responsiveness, improvements — is a contract renewal. Staggering by segment, giving notice in advance, offering options (terms, volumes, service versions) turns a unilateral announcement into a negotiation with a frame.

Third: accept that losing certain customers is the system working. The company that never loses a customer to price is, with mathematical certainty, undercharging almost everyone. The goal of a pricing policy is not universal retention: it is that every commercial relationship be knowingly profitable, and that exceptions be decisions — not year-end discoveries.

Price is already telling a story in every quote that went out this week. The only open question is who writes it.