Raise your price for new clients immediately, and move existing ones at a natural break: a new project, a new year, a new phase. Tell them once, in one sentence, with a date. No apology and no essay. Most people say fine, and the ones who leave were the ones costing you the work you actually wanted.
The hard part is not the conversation. It is deciding the number and then not flinching when you say it.
First, check the number is real
Before you raise anything, work out what you actually need. Most people who feel underpaid have never done the arithmetic, so they negotiate from a feeling rather than a floor.
Take the income you need, add tax and costs on top, then divide by the hours you can honestly bill. Not 40 a week. Between sales, scoping, invoicing, admin and the weeks nothing is booked, 20 to 25 is a realistic ceiling for a solo freelancer, and hitting it consistently is a good year.
If the result is far above what you charge today, that is the finding, not an error in the maths.
Raise it for new clients first
This is the low-risk half and almost nobody does it deliberately. The next person who asks gets the new number. Nothing to announce, nobody to negotiate with, no relationship at stake.
Do this for a few weeks before touching anyone existing. Two things happen. You find out whether the market accepts the number, which removes most of the fear. And you build a reference point: when you do talk to an existing client, you are describing your current rate rather than proposing an increase.
Move existing clients at a natural break
Mid-project is the worst moment. You are asking someone to renegotiate something they already agreed, while depending on you to finish it. It reads as leverage even when it is not.
Natural breaks:
- A new project or a new phase of the same one
- The start of a year or a quarter
- A retainer renewal
- Any point where the scope changes anyway
At one of those, say it once:
From September my day rate is 700. Everything we have already agreed stays at the current rate.
That is the whole message. No justification, no market conditions, no paragraph about rising costs. Explaining invites negotiation of the explanation. A rate is a fact about you, not a case you have to win.
Raise it when you are busy, not when you are desperate
A queue is the only real leverage in this job.
If you are turning work down, your price is too low and raising it costs you nothing: the work you lose is the work you did not have room for anyway. If nothing is coming in, fix the pipeline before you touch the price. Cutting it in a slow month makes the phone ring with exactly the clients you do not want, and those clients are harder to get rid of than they were to get.
What to say when someone pushes back
Four responses, and which to use when.
"That is more than we budgeted." Ask what the budget is. If it is close, either meet it once or remove scope to fit. If it is far off, say so plainly and part well. A client whose budget is half your rate is not a negotiation, it is a mismatch.
"Can you do it for the old rate just this once?" Sometimes yes, if the relationship is worth it and you say explicitly that it is a one-off. Say it out loud, because an unstated exception becomes the new rate.
"Why has it gone up?" "My rate reflects where my work is now." Then stop talking. The silence does more than another sentence would.
Silence. Some clients simply do not reply, and then come back three months later at the new rate. Do not chase and do not soften the number in a follow-up.
Have a floor, written down
Decide the number below which you say no, before anyone asks. Write it somewhere you will see it.
Without one you get the slow drift: a small favour for an existing client turns into a fortnight, at a rate you agreed to in a hurry. Cheap work does not stay cheap. It takes the same meetings, the same emails and the same bug reports as expensive work, and it crowds out the client who would have paid properly, because you were busy.
What actually happens
Almost everyone who raises their rate reports the same two things. Fewer clients said no than they expected, and the ones who left were the difficult ones. That is not a coincidence: price is a filter, and the clients most sensitive to it are usually the ones most expensive to serve.
The version of this that goes wrong is the one done quietly and inconsistently: a different number for every client, no floor, and an apology attached to each one. Pick the number, say it the same way every time, and let it do the filtering.