Raise your rates by telling existing clients in a short written notice, about 30 days ahead, stating the new price and the date it starts, without an apology and without a justification. The single most common mistake is not the size of the increase. It is turning a business decision into a request for permission, which invites a negotiation that was never on the table.
The second most common mistake is raising everybody at once. Rate increases carry real risk, and there is no reason to take all of it in the same week when you can sequence it and learn from the first few.
When should you raise your rates?
When at least one of three things is true: your calendar is full at the current price, your costs have moved, or your work is demonstrably better than it was when the price was set. Full capacity is the clearest signal, because a business turning down work at a given price has proved that price is below market.
The trap is waiting for a feeling of deserving it. That feeling arrives late or never, and it is not the relevant test. The relevant test is whether the current price still reflects what the work is worth to the buyer and what it costs you to deliver, and both of those move over time whether or not you feel ready.
The other trigger worth naming is resentment. If you notice that a particular client's work has started to feel like an imposition, that is information. It usually means the price stopped matching the effort some time ago, and the mismatch has been showing up as an emotion instead of as a number. Repricing fixes the emotion, and nothing else does.
How much notice should you give existing clients?
Around 30 days, and no more than 60. Long enough that nobody feels ambushed, short enough that it stays a small decision rather than becoming a scheduled review with other people involved.
The failure at the short end is obvious: an increase applied to work already underway or announced days before an invoice reads as opportunistic, and it damages trust out of all proportion to the money. The failure at the long end is less obvious and more common. Ninety days is enough time for the client to add it to a budget cycle, put it in front of a manager, and turn a routine change into a procurement exercise with a comparison against alternatives. You do not want a comparison. You want a small adjustment that gets approved by the person you already deal with.
For anything with a defined term, the increase belongs at the renewal boundary rather than mid-term, both because it is cleaner and because the renewal already exists as a natural point for terms to change. Mid-term increases on a fixed-price project are a different thing entirely and usually indicate the scope moved, which is a scope conversation rather than a pricing one.
What should the message actually say?
Three sentences. That the rate is changing, what it will be, and the date it takes effect. Optionally a fourth line thanking them for the work. Nothing else.
The instinct is to explain, and the explanation is what causes the trouble. Naming rising costs invites a discussion of your costs, which are not the client's concern and not a strong argument. Citing your increased experience invites the observation that they were happy with the previous standard. Apologising signals that you regard the increase as unreasonable, and the client will take your word for it. Every additional sentence is a surface to push against.
Short and neutral works because it matches the form of every other supplier price change the client receives, all of which are absorbed without comment. Their software costs more this year. Their insurance costs more. None of those arrive with an apology, and none of them get negotiated. Presenting yours in the same register categorises it correctly before anyone has read the number.
One practical note: send it in writing, on its own, not attached to an invoice and not raised at the end of a call about something else. It needs a subject line and a date, because it is a term change and both of you may want to find it later.
Who should you raise first?
New clients, immediately, and then work backwards through your existing list in order of how little you have to lose. The order is the whole risk management strategy here.
Every new prospect from today gets quoted the new rate. There is no history, no comparison and no conversation, and within a couple of months you have real evidence about whether the market accepts it. That evidence is worth far more than any amount of deliberation, and it costs nothing to gather.
Then the existing list, in three passes. The quiet, profitable, low-maintenance clients first, because the risk is genuinely small and each acceptance builds your nerve for the next. Then the middle. Then the largest and longest-standing accounts last, because those are the relationships with the most history, the most complicated internal approval, and the most to lose, and by the time you reach them you will have run the conversation several times and stopped dreading it.
What if a client says no?
Then you have a decision that was already made for you by arithmetic, and you should make it calmly. Either the relationship is worth keeping at the old price for a defined period, or it is not, and the honest way to answer is to look at what that client actually costs you to serve.
There is a middle path worth knowing about, and it is not a climbdown. Hold the rate and reduce the scope. "I can keep the current price for what we're doing, and here is what would come out of it" preserves your pricing while giving them a genuine choice. Clients frequently take the reduced scope, and the ones who do were telling the truth about the budget.
What is worth avoiding is the quiet exception. Reversing the increase for one client and not the others creates a rate you cannot explain and a precedent that resurfaces at every future review, and it tends to become known. If you are going to grandfather someone, put a date on it, say so once, and hold the date.
And expect some attrition. A small amount is a sign the increase was real, not a sign it failed. The clients who leave over a modest, well-noticed increase are, with striking consistency, the ones who negotiated hardest at the start, paid slowest, and consumed the most unbilled time. Losing them is the mechanism working, not the mechanism breaking.
Is it easier to raise prices or replace clients?
Raising is cheaper, and it is not close, which is the argument most people miss while worrying about attrition.
Winning a replacement client costs research, outreach, a proposal, a call, and weeks of calendar time, all unpaid. Raising an existing client's rate costs one email. If a ten percent increase across ten clients loses you one, you are ahead on revenue, ahead on time, and ahead on capacity, and you found out something useful about the client you lost.
That comparison is also the reason to treat this as a routine annual event rather than a rare crisis. A business that reviews its rates every year makes small, unremarkable adjustments that clients absorb. A business that avoids it for four years eventually needs a large correction, which is far more likely to be refused, and which arrives with an unspoken admission that the price was wrong for a long time. The infrequency is what makes it frightening, and the fear is what causes the infrequency.
The honest hard part
The hard part is that this is not a pricing problem, it is a self-perception problem wearing a pricing problem's clothes, and no amount of scripting fixes the underlying thing.
What people are actually afraid of is not the lost revenue. It is a specific two-line reply that says the client has decided to go elsewhere, and the interpretation that follows it, which is that they were only ever tolerating you and the price was the thing holding it together. That fear is why the message gets softened into a question, why the explanation gets added, and why the whole thing gets postponed for another year.
The practical counter is to remove the discretion. Pick a date in the calendar, decide the new number in advance, write the three sentences, and send them on the date whether or not you feel ready. The feeling is not a signal about your work and it is not going to improve on its own. Every operator who has done this a few times reports the same thing: the great majority of clients reply with some version of "no problem," and the anticipation was worse than the entire event by a wide margin.
If the deeper issue is that the original number was set from insecurity rather than arithmetic, the fix belongs further upstream, and we set out how to derive one properly in how to price a service with no track record.
