Raising prices without losing the book

Two kinds of advice circulate about raising your prices, and both are wrong in the same way: they treat it as a feeling. The confidence camp says you are underpriced, you have always been underpriced, and the only thing standing between you and the number you deserve is nerve — send the email, hold your breath, the ones who leave weren't paying you enough to keep anyway. The churn-panic camp says the opposite with equal conviction: your best clients are your longest clients, long clients are loyal because the relationship works at the current number, and touching it risks the whole book for a few points of margin. Neither camp does arithmetic. Both are giving you a mood and calling it a strategy.

The mood is understandable — a price increase is one of the few business decisions where you get to watch, in real time and by name, exactly who objects — but it is also avoidable. You can calculate, before you send a single message, exactly how many clients an increase can afford to lose and still leave you ahead. You can decide, before anyone gets an email, which clients hear about it first and which hear about it last, and there are good reasons for that order that have nothing to do with who you like better. And you can write the message itself in a way that either invites a negotiation or closes the door on one, which is a choice, not an accident of how apologetic you happened to feel that morning. None of this makes the increase painless. It makes the pain proportional to something real instead of to how the last difficult call went.

The number that should set your ceiling

Start with the arithmetic, because it reframes the whole decision. A price increase does not need to retain every client to be worth doing — it needs to retain enough of them, and "enough" is a number you can compute from your own margin before you send anything.

Say you raise prices by 10 percent and your current gross margin on the work is 40 percent. The break-even churn rate — the share of clients you could lose entirely and still make the same total profit as before the increase — is roughly the increase divided by the sum of the increase and the margin: 10 ÷ (10 + 40) = 20 percent. You could lose one client in five and be no worse off than if you'd never touched the number. Anything better than that, and the increase made you money even accounting for the clients who walked.

The reason this matters is that it inverts the instinct most owners have. The instinct says: don't raise prices, you might lose clients. The arithmetic says: you can afford to lose a specific, calculable number of clients, and the real risk isn't losing some — it's losing more than the break-even line, which is a much narrower and more answerable question. Run the same math at a 5 percent increase with the same 40 percent margin and the break-even churn drops to about 11 percent — a smaller increase buys you less room for people to leave, which is worth knowing before you decide a "modest, safe" increase is automatically the lower-risk move. It often isn't; it just fails more slowly.

The other side of this arithmetic is margin-specific, and it's worth saying plainly: a company running thin margins has almost no room here. At a 15 percent margin, a 10 percent increase only breaks even at roughly 40 percent churn — a share few businesses should plan around losing. Thin-margin firms need bigger increases or fewer of them, and the number tells you that before a single client has to. This is also where competing on price quietly punishes you twice: a thin margin was already the cost of that position, and it's the same thin margin that now leaves almost no room to raise anything.

Who hears about it first, and why the order isn't arbitrary

The sequence that works, and that most owners get backwards under pressure, is new prospects first, then clients at their renewal or contract boundary, and existing mid-engagement clients last, if at all before their natural break point.

New business is the easiest call in the entire exercise because there is no relationship to protect and no history to explain — a prospect who hasn't signed yet simply meets the new number as the number, with nothing to compare it to. Every day you delay updating the rate you quote new prospects is a day you are selling the old, wrong number to people who have no idea it changed, which is the cheapest kind of margin to leave on the table because fixing it costs nothing but updating a document.

Renewals come next, and they are close to free in a different way: a renewal is already a moment where the client expects to think about the relationship and the terms, so a new number arriving with the renewal reads as the natural shape of things rather than an interruption. This is also where hourly, fixed-fee and retainer structures matter, because the ease of this step depends heavily on which one you're running — a retainer with a defined term has a built-in renewal conversation already scheduled, while true hourly work with no term at all has no natural moment to hang an increase on, which is exactly why hourly engagements tend to drift years past the point where the rate should have moved.

Legacy accounts — long-standing clients mid-contract, with no renewal date in sight — are the ones that require actual judgment, and they're also where the confidence-camp advice does the most damage by treating them the same as everyone else. These are usually your most profitable relationships precisely because the price has been static while your costs and skill have not, and they are also the relationships where a clumsy increase does the most reputational harm, because clumsiness reads as a change in how you regard the relationship, not just a change in a number. The right move here is rarely to force the increase mid-term. It's to wait for the next natural boundary — end of project, end of fiscal year, contract renewal — and treat that boundary as the trigger, the same way it is for everyone else. Forcing it earlier because you're impatient converts a scheduling decision into a trust decision, and trust decisions are the expensive kind.

What a fair runway looks like

Notice period should scale with tenure and with how much the client's own planning depends on your number, not with how awkward the conversation feels to have. A client of a few months has little standing to expect a long runway — thirty days' notice, delivered cleanly, is fair and sufficient. A client of two or three years, whose own budgeting has quietly assumed your rate as a fixed line item for a long time, has earned more than that: a full quarter is the reasonable floor, and longer is better if the engagement is large relative to their overall spend. The test worth applying is whether the client could plausibly need to requote their own downstream work, adjust their own pricing, or find a replacement vendor in the time you've given them — if the honest answer is no, the notice period is too short regardless of what your contract technically permits.

Why the justification paragraph backfires

The instinct when writing the increase message is to explain — costs have gone up, the market has moved, the team has grown, here is the reasoning that makes this fair. Almost every version of that instinct makes the message worse, because a paragraph of justification does not read as fairness. It reads as an opening position in a negotiation, and clients who were never going to push back suddenly have a list of specific claims to push back against. If you tell someone their rate is going up because your costs rose 8 percent, you have just handed them a lever: they can ask for the receipts, propose splitting the difference, or point out that their own costs haven't risen and ask why that's their problem. None of those conversations existed before the justification did.

The messages that hold are short, stated as fact rather than argued as fair, and specific about the number and the date: the new rate, when it takes effect, and what stays the same. "Starting with your March renewal, the rate for this engagement will be $X, up from $Y. Everything else about the engagement is unchanged." That's the whole message. It closes the door on negotiation not by being cold but by not opening a door in the first place — there is nothing in the sentence to argue with, because it isn't an argument.

The client who leaves and comes back

Some clients will leave over the increase, and a fraction of those will eventually ask to come back — sometimes months later, sometimes after trying a cheaper option and getting what they paid for. This is the moment where all the discipline of the increase gets tested, because the easy, relationship-preserving move is to let them back in at their old rate, and it is also the move that undoes the entire point of having raised prices in the first place.

Let them back in at the current number, not the number they left at. Anything softer teaches two lessons at once, both bad: it tells the client who left that objecting hard enough eventually works, which trains your best negotiators to leave every time you touch pricing, and it tells you nothing true about whether the increase was correctly priced, because a returning client at the old rate isn't evidence the increase worked — it's evidence you didn't actually hold it. The break-even arithmetic already accounted for a share of clients leaving. A returning client at the current rate is simply that arithmetic playing out one relationship later than expected, and it's a fine outcome. A returning client at a discount is the arithmetic quietly failing while looking, on the surface, like a happy ending.