30 July 2026 ยท 8 minute read
The price rise that looks free
Margin arrives in the first month and the bill arrives over the following year. That is why the spreadsheet and the outcome so rarely agree.
Every operator has seen this one. The price has not moved in three years, costs have, and somebody finally says the obvious thing. Twenty percent. The model is built in an afternoon: current revenue, times 1.2, minus an allowance for the customers who will leave. The allowance is a round number that somebody in the room felt comfortable with.
The answer comes out large and positive, which it will, because the structure of the calculation guarantees it. Revenue moves immediately. Cost does not move at all. The only thing standing between the company and free money is the allowance, and the allowance was a feeling.
Then the year happens, and the year does not look like the afternoon.
What the afternoon left out
Three things, in the order they arrive.
- New business stops first. A prospect with no switching cost to pay is the most price sensitive person in your market, and they see the new price immediately. The pipeline thins in week two and nobody notices because the pipeline always moves
- Competitors answer, partially, and late. Not the same month, and not by the full amount. They see the gap open, they take a few months to decide, and then some of them close part of it. Your relative position ends up better than before and worse than the spreadsheet assumed
- Existing customers respond at their renewal dates and not before. This is the one that ruins the timing of the whole plan, because it means the damage lands at whatever moment your contract book decides, which has nothing to do with your financial year
The first two are annoying. The third is the one that makes the spreadsheet actively misleading, because it puts the revenue benefit in months one to six and the customer loss in months seven to eighteen, and then somebody declares victory at the half year.
The shape of it
Profit against price, computed rather than drawn
A worked example, on an invented company
Harborline Components is invented. It is a contract manufacturer with 24 customers, 14.24 million dollars of annual revenue, 31 percent gross margin, 46 people, a largest customer at 18 percent of revenue, a top five at 54 percent, one single sourced marine grade billet, and a price it has not moved since 2023. It exists because these questions get asked about companies of exactly that shape.
Run 20 percent across its twin and the first six months look like the afternoon predicted. Margin is up, cash is up, and nothing has visibly broken. The customer list has long contracts on it and most of those contracts have not come up yet.
Then the renewals start landing. One large account reaches its date, compares what it pays against what is available, and goes out to tender. Two others renew and ask for something in return, which costs either price or scope. Meanwhile the pipeline has been thin since month two, so there is nothing arriving to replace what leaves.
The twelve month figure is still positive. The twenty four month picture is a different company with fewer, larger, more price aware customers, which is a strategic position somebody should have chosen deliberately rather than arrived at by arithmetic.
The same move, two accounts of it
| Month | What the afternoon model says | What has to happen first |
|---|---|---|
| 1 | Revenue up twenty percent | Revenue up on the portion of the book not under contract. Everyone else pays the old price until their term ends |
| 2 | Same | Win rates drop on new business. Nobody attributes it to the price change yet because pipelines are noisy |
| 4 | Same | A competitor decides to move and schedules part of a match |
| 6 | Half year looks excellent | It does look excellent. This is the month the decision gets declared a success |
| 8 | Same | The first large renewal lands. The account either leaves or renews having extracted something |
| 12 | Annualised benefit banked | Benefit is real and smaller. Customer count is down, the remaining base is more concentrated, and the pipeline has not recovered |
| 18 | Not modelled | This is where the decision is settled, and nobody built a model that went this far |
Neither column is a forecast. The left one is what a multiplication produces. The right one is the sequence of things that have to occur for the left one to be either true or false, which is the part worth arguing about.
The thing to measure
How much of your revenue can move this year
Before arguing about elasticity, get the contract book out. Work out what share of revenue reaches a renewal date inside the horizon you are planning over. That number governs everything about the timing of the answer, and most companies can produce it in an hour.
A business with nine month terms is a different animal from one with three year terms, at identical elasticity. One finds out within the year. The other spends the year feeling clever and finds out in the next one.
The second thing to measure is your win rate on new business, weekly, starting the month before the change. It is the earliest honest signal you have, it moves before anything else does, and it is the cheapest possible moment to discover that the market is more price sensitive than you assumed.
- Share of revenue reaching renewal inside the horizon
- Win rate on new business, weekly, from the month before
- The date each of your five largest accounts next gets to make a decision
None of this is an argument against raising price
Most companies that have not moved their price in three years should move it. The margin is real, the cost pressure is real, and a business that will not charge for what it does eventually cannot afford to do it well, which is its own slow way of losing customers.
The argument is against the afternoon model. It produces a number that is right about the direction and wrong about the shape, and being right about the direction is the most dangerous kind of wrong, because it means the plan survives the first six months of evidence.
A price rise is not a revenue event. It is a sequence of decisions made by other people, spread across your contract book, at times you do not control. If your model does not have those people in it with their dates attached, what you have is a multiplication with an opinion inside it.
Run the question instead of arguing about it
Build a twin from whatever record you have, and see the range rather than the one number somebody liked.