How much should I raise my freelance rates by?

Enough to be worth the conversation. A rise of x lets you lose x/(1+x) of your book and still earn the same, so 20% tolerates losing 16.7% of your clients while 5% tolerates only 4.8%. Both cost you the same one awkward email, so a token 5% is usually the wrong trade — it buys a quarter of the money and a third of the margin for error. Below 10% you are barely keeping pace with inflation, which is not a raise at all.

How much notice should I give a client before raising my rate?

Thirty days is the minimum and sixty is the professional standard for ongoing retainers, because clients with budget cycles need a quarter boundary to move money. Anchor the new rate to a natural edge — a contract renewal, the start of a quarter, or the beginning of the next project — rather than an arbitrary calendar date. Never change the rate mid-project or mid-milestone; finish what was quoted at the price it was quoted at.

Should I raise rates for all my clients at the same time?

Stagger them, and start with the client you would be most relieved to lose. Your first version of this email is your worst version, so send it where the downside is an outcome you would have chosen anyway. Once one client has accepted, you write the next one from evidence instead of hope — and the difference in tone is visible on the page.

Should I grandfather existing clients at the old rate?

Only with an expiry date attached. Open-ended grandfathering guarantees that your longest-standing client is permanently your worst-paid one, and the gap widens every year you repeat it. Say that the current rate holds through a stated date and moves after it, rather than that the client is exempt. Same concession, finite cost.

FERNMILL

How to raise your rates with existing clients

Q: What do I say when a client asks why my rates are going up?

As little as possible, and nothing about your own costs. Every reason you supply is a surface the client can argue with: your rent going up is not their problem, and your growing experience is something they can dispute. One neutral sentence — that you review rates annually and this is the outcome — is complete. A price is a fact you state, not a case you prosecute.

Q: What if the client says no and leaves?

Work out beforehand how many you can afford to lose, so that a departure is a result rather than a shock. Offer a reduced scope at the old total before you offer a lower rate — cutting the deliverable protects your price, cutting the price resets it permanently. If they still leave, hand over cleanly with a stated end date and a short summary; the client who left over 20% is often the one who comes back at the new rate later.

The reason this conversation feels dangerous is that you are imagining the downside and not calculating it. Calculate it and the fear mostly evaporates, because the number is much friendlier than it feels: at a 20% rise you can lose one client in six and be exactly where you started, with the hours back.

The number that makes this easy

Raise your rate by x, and you break even after losing x ÷ (1 + x) of your billings. That is the whole calculation, and almost nobody does it before writing the email.

Rise	You can lose…	…and still earn the same
5%	4.8% of your book	with 4.8% of your hours back
10%	9.1%	9.1% of your hours back
20%	16.7%	16.7% of your hours back
30%	23.1%	23.1% of your hours back
50%	33.3%	a third of your hours back

Make it concrete. You bill 100 hours a month at $85 — $8,500 a month, $102,000 a year. You move to $102. Breakeven is $8,500 ÷ $102 = 83.3 hours, so two clients worth 16.7 hours a month can walk out and your income does not move an inch. You just work 16.7 fewer hours for it.

And if nobody leaves — which is the usual outcome — 100 × $102 = $10,200 a month. That is $20,400 a year from one email you were putting off.

Do this part first. "Raise by 20%" is only the right answer if you know what your hour actually has to earn once tax, unpaid admin and a realistic billable ratio are in the arithmetic. The free rate calculator runs that version — it usually reveals that the target rate is higher than the one people talk themselves into asking for.

Why a 5% rise is usually the wrong move

The standard advice is to nudge rates up a little each year. The problem is that a small rise and a large one cost you exactly the same thing: one uncomfortable email, one client's undivided attention, and roughly one year's worth of goodwill for asking. You can only spend that about once every twelve months.

So compare the trades. On $102,000 of billings, a 5% rise earns $5,100 and tolerates losing 4.8% of your book. A 20% rise earns $20,400 and tolerates losing 16.7%. Four times the money, three and a half times the margin for error, identical social cost.

There is a floor underneath this too. Hold $85 flat for three years while prices rise 3% a year and, in year-three money, your rate is worth $77.79 — an 8.5% pay cut you never agreed to. Standing still would have required $92.88. The first 9% of a three-year "raise" is not a raise. It is back pay you are not getting.

Send it to the client you can afford to lose — first

Everyone announces to their whole roster on the same day, which means the version of the email you have never tested goes to your best client simultaneously with everyone else. Do it in reverse.

List your clients, sort by rate ascending and friction descending, and send to the one at the top of that list — the underpaying, slow-replying, scope-creeping one you would be quietly relieved to see go. Three things follow:

The downside is an outcome you wanted. If they leave, you have freed capacity from your worst-paid work at the exact moment you have a higher price to sell it at.
Your first attempt is your worst attempt. The wording, the objections, your own nerve — all of it improves after one real run. Spend the bad draft where it is cheap.
The second email is written from evidence. "My rate moves to $102 from 1 October" reads differently when someone has already agreed to it, and clients can hear the difference even though nothing on the page has changed.

The announcement email

Sixty days for anything ongoing — clients with budget cycles need a quarter boundary to move money, and thirty days often lands after the budget is set. Anchor the change to a natural edge: a renewal, a quarter start, the next project.

Subject: Rate change from 1 October Hi Dana, A heads-up well in advance: from 1 October my rate moves from $85 to $102 an hour. I review rates once a year and this is where it lands for 2027. Everything already scheduled or in progress stays at $85 — the November campaign work is quoted and that quote holds. The new rate applies to work commissioned from 1 October. Nothing else changes: same availability, same turnaround, same day-to-day. Happy to talk it through any time before then. Sam

Why it works. One sentence of context, and it is about your process rather than your costs. Your rent is not the client's problem and your growing experience is something they can dispute; "I review rates annually" is neither. Note the second paragraph — protecting quoted work is what turns this from a squeeze into a policy, and it costs you nothing, because that work was already sold.

No "because" beyond that. Every reason you volunteer is a surface to argue with. Freelancers lose this negotiation in the third paragraph, explaining, far more often than they lose it in the first.

The client with no ongoing contract

If there is no retainer and no live project, do not send an announcement at all. There is no agreement to amend, so a rate-change email invents a negotiation that did not exist and invites a response to a question nobody asked. Put the new number in the next quote and say nothing.

Subject: Re: new landing page — quote Hi Dana, Good to hear from you. Scoped it as: Landing page, 2 rounds of revisions — 14 hrs @ $102/hr, $1,428 Delivery — 12 working days from sign-off Happy to start the week of the 14th if that works. Sam

If they notice and ask, answer in one line: "Yes — rates moved in October." That is the entire conversation, and it is a conversation you win by not having prepared a speech for it.

The two kinds of pushback

Pushback 1
"Can you hold the old rate for us?"

Cut the scope, never the rate. A discount resets your price permanently and every future quote is negotiated against it. A smaller deliverable at the old total leaves the price intact and hands the decision to the client.

Subject: Re: rate change Hi Dana, Understood — let us keep you at the same monthly spend instead. Right now the retainer is 24 hrs at $85 = $2,040/mo. At $102 that same $2,040 buys 20 hrs, so: Option A — $2,040/mo, 20 hrs. Budget unchanged. We drop the weekly written report and I batch the small requests into one Thursday slot. Option B — $2,448/mo, 24 hrs. Everything exactly as it is now. Either is fine by me. Tell me which and I will set it up from 1 October. Sam

Why it works. The rate is not on the table in either option, so it never gets argued about — both options price the hour at $102, and the client can check that. They are choosing between two things they can have rather than being refused the one thing they asked for. And if they pick A, you have recovered four hours a month without touching your headline price.

Pushback 2
"We can't stretch to that."

Sometimes true. Leave warmly and on a date, because the client who could not afford 20% this year is a plausible client at the new rate next year — and a hostile exit forecloses that.

Subject: Re: rate change — wrapping up Hi Dana, No problem at all, and thank you for saying so directly. Let us finish cleanly: I will complete everything in the current sprint and hand over on 30 September. I will send across the working files, the brand asset folder and a short note on where each thing lives, so whoever picks it up is not starting cold. It has been a genuinely good three years. If budgets look different in 2027, come back to me — I will always find room for you. Sam

Why it works. A clean handover is cheap to give and disproportionately valuable to receive; it is also the thing people mention when they refer you on. The last client you lost well is a better referral source than most of the ones you kept.

When not to do this

Mid-project or mid-milestone. Finish what was quoted at the price it was quoted at. Changing the number after work has started is the move that makes people distrust everything else you send them.
Right after something went wrong. Missed deadline, botched delivery, escalation. Wait for a clean stretch — the raise reads as an argument about value, and you want the recent evidence to be on your side.
While an invoice is overdue. Get the account current first; two asks at once and neither gets resolved. The follow-up sequence for that is here.
Less than twelve months since the last one. Annual is a policy. Twice a year is a pattern nobody can budget around.

One thing that is not a reason to wait: the client whose scope has been quietly expanding. That is not a rate conversation at all — it is a scope conversation, and merging the two lets a change you should be charging for disappear into a percentage. And if your contract has an auto-renew or a fixed-rate clause buried in it, the notice you owe may be contractual rather than courteous — the free contract checker flags both in about thirty seconds.

Common questions

How much should I raise my rates by?

Enough to be worth the conversation. A rise of x tolerates losing x/(1+x) of your book, so 20% tolerates 16.7% while 5% tolerates 4.8% — for the same one awkward email. Below 10% you are barely keeping pace with inflation, which is not a raise.

How much notice should I give?

Thirty days minimum, sixty for ongoing retainers, anchored to a renewal or a quarter start rather than an arbitrary date. Never mid-project.

What do I say when they ask why?

One neutral sentence about reviewing rates annually, and nothing about your costs. Every reason you supply is something the client can argue with. A price is stated, not justified.

Should I raise everyone at once?

Stagger, starting with the client you would be most relieved to lose. Your first attempt is your worst attempt — spend it where the downside is an outcome you wanted anyway.

Should I grandfather existing clients?

Only with an expiry date. Open-ended grandfathering makes your longest-standing client permanently your worst-paid one, and the gap grows every year you repeat it.

What if they say no and leave?

Offer reduced scope at the old total before you offer a lower rate — cutting the deliverable protects your price, cutting the price resets it. If they still go, hand over cleanly and on a stated date.

Work out the number before you write the email

The rate-calculator skill computes the floor your hour actually has to clear — tax, unpaid admin, holiday, a realistic billable ratio — and then shows the annual gap between that floor and what you currently charge. It shows every line of the arithmetic so you can check it, which is the point: the reason most rate rises are too timid is that the person asking has not seen the real number. It is free and MIT licensed.

Get it free on GitHub All 14 skills — $39