How to Raise Your Agency Rates Without Losing Clients

Almost every agency is underpriced, and the ones most likely to be underpriced are the ones whose clients are happiest.

That is not a coincidence. Nobody complains, so nothing prompts a review. Costs rise quietly - salaries, software, insurance, rent - while the rate stays where it was set three years ago. The gap compounds annually, invisibly, and the first visible symptom is not a client complaint. It is that the team is fully booked and the business still cannot afford to hire.

This guide covers when to raise rates, how much, the sequencing that minimises churn, the exact words that work, what to do when a client pushes back, and why the delay is more expensive than the increase.

The case for raising, stated plainly

Three arguments, in ascending order of how often they apply.

Your costs have risen. This is the least interesting reason and the most defensible. If salaries went up 5% and your rate did not, your margin fell by roughly that much. A rate held flat is a real-terms price cut, and holding it for three years is a compounding one.

You are better than you were. An agency two years further into a specialism delivers more, faster, with fewer mistakes. Under hourly billing, getting faster actively reduces your revenue - the perverse incentive covered in our guide to pricing agency services. A rate increase is partly how you stop being punished for competence.

Your price is filtering wrongly. This is the one most agencies miss. Price is a signal, and a price materially below the market tells sophisticated buyers something - usually that you are junior, or desperate, or about to be overwhelmed. Agencies that raise rates frequently report that the quality of inbound enquiry improves, not just the value. You stop attracting the buyers for whom price is the only variable, and those are the buyers who generate the most scope disputes and the least profit.

When to do it

Do not look for the right moment. Create one.

Set a fixed annual review date and put it in the calendar - the same month every year, regardless of how things feel. A scheduled review removes the need to work up the nerve, which is the actual obstacle. Most agencies that "cannot find the right time" have a nerve problem rather than a timing problem, and a date in the diary solves it.

Beyond the annual review, four situations justify an out-of-cycle increase:

You are consistently at capacity. If you are turning work away or quoting six weeks out, demand exceeds supply at your current price. That is the textbook signal, and it is the least risky moment to move.

You have won something that changes your standing. A recognisable client, a strong measurable result, a specialism that is now demonstrably yours.

Your costs jumped. A significant salary correction, a large software increase, a move.

You dread specific accounts. Not always a pricing problem - sometimes it is a boundaries problem - but a client you resent serving is very often one you underpriced and have been quietly subsidising ever since.

How much

Two separate questions, and conflating them is why increases get stuck.

For new clients: go to the right number

New prospects have no reference price. They hear your rate as the rate, not as an increase. So do not creep - move to where you should be.

Derive it from your real cost floor rather than from what feels acceptable. The calculation is in our pricing guide, and the part everyone gets wrong is the denominator: dividing costs by 40 hours a week rather than genuine available hours produces a break-even figure far below reality. Asana's guide to utilization rate and Scoro's breakdown of billable utilization both explain why available hours, not total hours, is the correct basis - and our own post on utilization rate covers the competing formulas.

If the honest number is 40% above what you charge now, charge it to the next new prospect. The worst outcome is that they decline, which costs you a deal you were probably going to lose money on.

For existing clients: increase in steps

Existing clients have a reference price, so the increase is felt as a change rather than a fact. Here, gradualism is genuinely the better strategy.

5-15% at a renewal is normal and rarely contested. Above 20% needs a specific justification - a scope change, a materially different service, or an increase that has been signalled in advance.

If an existing client is dramatically underpriced - say 50% below your new rate - do it over two increases twelve months apart, and say so at the first one. "We're moving to £X now and £Y next year" is far better received than an unannounced 50% jump, and it lets them plan.

The sequencing that minimises churn

The order matters more than the size.

1. Raise new-client rates immediately. No notice required, no conversation, no risk to existing revenue. Do this first, always. You will also learn quickly whether the market accepts the new number, which de-risks step three.

2. Raise at natural boundaries for existing clients. A retainer renewal, the end of a project, the start of a new engagement. A rate change at a natural boundary is expected. The same change mid-engagement feels like a renegotiation of an agreement already made - and on a fixed-fee project it usually is one, so do not do it.

3. Give 60 days' notice. Enough for the client to budget, escalate internally, or decide. Thirty days is acceptable; anything less reads as an ultimatum. Never apply an increase retroactively or with a single billing cycle's notice.

4. Tell everyone in the same week. Clients talk, particularly within an industry. Finding out from a peer that they got a better deal is corrosive in a way the increase itself is not.

The conversation

Written first, then a call for your larger accounts. The written notice does the work; the call handles the reaction.

What to write

Four elements, in this order, kept short:

The change. From 1 January our rate moves from £X to £Y.

The reason. One sentence, true, without over-explaining. "Our costs have risen and this is our first increase since 2023."

What stays the same. The team, the scope, the way you work.

The date. Fixed and unambiguous.

Four sentences, in plain language, sent by a named person rather than an accounts address.

What not to write

Do not apologise. "I'm really sorry to have to do this" tells the client you think the increase is unjustified, which invites them to agree. You are not doing something to them; you are updating a price, as every supplier they have does.

Do not over-justify. One reason is credible. Four reasons read as a defence, and a defence implies you expect an attack. The longer the explanation, the weaker it sounds.

Do not present it as negotiable. "Let me know if this is a problem" is an invitation, and someone will accept it. If you are genuinely willing to make exceptions, decide in advance which ones and why - do not discover it mid-conversation.

Do not blame the client. "The scope has grown" may be true, but it belongs in a separate scope conversation with a change order, not bundled into a rate increase. Mixing the two makes both harder to resolve.

The call, for your larger accounts

Anyone above a meaningful revenue threshold gets a call before the email, not after. It is a courtesy, and it lets you handle the reaction directly rather than reading it in a reply drafted in frustration.

Keep it to three minutes: the change, the reason, the date, and then stop talking. The silence after is uncomfortable and it is where the client processes it. Filling that silence with additional justification is the most common way these calls go wrong.

When they push back

Most do not. Of those who do, the responses fall into four types.

"That's a big jump." Acknowledge and hold. "It is - it's our first increase in three years, so it's catching up rather than getting ahead." Then stop. The most common failure here is answering an observation as though it were an objection.

"We can't afford it." Sometimes true. The response is scope, not price: "Understood. We could stay within your current budget by reducing the retainer to 30 hours - would that work?" This protects the rate, which is the thing you cannot get back, and gives them a real option.

"We'll have to look at alternatives." Reasonable, and often a negotiating position. "That's fair - I'd do the same. If it helps, I'm happy to talk through what's involved in a transition so you can compare properly." Confidence here is more persuasive than concession, and offering to help them evaluate signals you are not worried.

Genuine anger. Rare, and usually a symptom of something else - a delivery problem, a relationship that was already strained, a person under pressure internally. Do not resolve it in that conversation. "I can hear this has landed badly. Let's talk properly on Thursday" gives both sides room.

Expect to lose some, and know which

The fear is that clients will leave. Some will. The useful part is that you can predict which.

The clients who leave over a 10% increase are, with high consistency, the ones who were already the least profitable - the most price-sensitive, the most scope-flexible, the most demanding relative to what they pay. Losing them frees capacity for work at the new rate, and the arithmetic frequently improves immediately even before that capacity is refilled.

Two things to run before you send anything:

Calculate your break-even churn. If you raise rates 15%, you can lose roughly 13% of revenue and be no worse off. Knowing that number in advance turns a departure from a crisis into an expected outcome.

Rank your clients by margin, not revenue. Our guide to utilization and realisation covers why revenue alone misleads - a large account with heavy write-offs can be worth less than a small one that pays on scope. You will usually find one or two accounts you would genuinely be better off without, and knowing that beforehand makes the conversation much easier to hold.

Raising the rate on a retainer

Retainers need their own handling, because the client is not buying a rate - they are buying a monthly commitment, and the number they feel is the monthly figure.

Three approaches, in order of how well they are usually received:

Increase the fee, hold the allocation. The clean version. "From January the retainer moves from £4,000 to £4,600 for the same 40 hours." Transparent, easy to compare, easy to object to. Best where the relationship is strong and the work is clearly valuable.

Hold the fee, reduce the allocation. "From January the retainer stays at £4,000 and covers 34 hours." Mathematically identical and psychologically very different - the monthly cost does not change, which for a client managing a budget line is often the binding constraint. Some find it more palatable; others find it slippery, so read the relationship.

Increase the fee and add something. "The retainer moves to £4,600 and now includes the monthly reporting pack and quarterly strategy session." Easiest to accept, and only honest if the additions are real and were genuinely not included before. Bundling in something you were already doing for free is a discount you have now made visible, and clients notice.

Whichever you pick, do it at renewal rather than mid-term, and give the full 60 days. A retainer increase applied at the next billing cycle feels like a unilateral change to a running agreement, because that is what it is. Our guide to retainer models covers the structure that makes these reviews routine rather than exceptional.

If you genuinely cannot raise rates yet

Sometimes the honest answer is that the market will not currently bear your target rate. Three things worth doing before concluding that.

Check that you are comparing like with like. Agencies frequently benchmark against firms with a different cost base, a different specialism, or a different client size. A generalist competing with specialists on price will always lose that comparison; the fix is positioning, not pricing.

Raise the floor rather than the rate. Introduce a minimum engagement size. This is often more effective than a rate increase because it removes the small, fiddly projects that consume disproportionate coordination time - the ones where the rate looks fine and the margin is terrible.

Fix realisation before rate. If you are writing off 15% of tracked billable time at invoicing, recovering that is equivalent to a 15% rate increase with no client conversation at all. Our post on utilization and realisation covers how to measure the gap; for many agencies it is the larger and easier win.

The signal you are already too cheap

Four indicators, any one of which means the market would bear more than you charge.

You win nearly everything you quote for. A win rate above roughly 60% usually means you are the safe cheap option rather than the considered choice. Losing a third of proposals on price is healthy.

Nobody ever questions the price. Occasional pushback is a sign you are near the edge of what the market will pay, which is where you want to be. Universal easy acceptance means you are well inside it.

You are booked out weeks ahead. Demand exceeding supply is the textbook signal, and agencies routinely respond by working longer rather than charging more.

Clients tell you that you are good value. Meant kindly and worth hearing precisely. "Good value" means cheap relative to the quality received, which is a compliment about your delivery and a comment on your pricing.

What not to do

Four moves that reliably make an increase worse.

Announcing it and then negotiating individually. Once one client discovers another got an exception, the increase is no longer a policy, it is an opening bid. Decide your exceptions in advance, apply them consistently, and do not explain one client's terms to another.

Bundling it with bad news. A rate increase in the same email as a delay, a staffing change or an apology invites the client to connect them. Separate messages, separate weeks.

Raising it mid-project on fixed-fee work. You agreed a price for a defined scope. Changing it mid-delivery is a breach of the thing that makes fixed-fee trustworthy, and no amount of cost justification repairs that. Wait for the boundary.

Softening it by adding unpaid work. "The rate goes up but we'll throw in the monthly report" converts a price increase into a scope increase, which is a worse deal than the one you had. If you want to add value, add it separately and later.

The cost of waiting

The delay is more expensive than the increase, and this is the part worth sitting with.

An agency billing £500,000 a year that postpones a 10% increase for twelve months does not lose £50,000 once. It loses £50,000 that year, and the following year's increase starts from the lower base, and the year after that too. The compounding runs against you for as long as you defer.

Meanwhile the internal cost accumulates. Underpricing shows up as an inability to hire, which shows up as an overloaded team, which shows up as delivery quality and eventually attrition. By the time the pricing problem is visible in the work, it has been a pricing problem for two years.

The agencies that hold rates flat for years are rarely doing so from analysis. They are doing so because raising prices is uncomfortable and there is always a reason to wait - a quiet quarter, a client mid-project, a market that feels uncertain.

Which is why the single most effective intervention is not a better script. It is a date in the calendar, reviewed whether or not it feels like the right year, so that the decision is made by a process rather than by nerve.

Communicating an increase to a long-standing client

The hardest version is the client you have had for five years at a rate set in year one. The relationship is genuinely warm, which makes it harder rather than easier.

Two things help.

Acknowledge the history honestly, once. "You've been with us since 2021 and we haven't changed our rate in that time, which is on us rather than on you." This is true, it is disarming, and it pre-empts the obvious objection - that the increase is sudden - by naming the reason it feels sudden.

Do not let warmth become a reason to stay underpriced. The most common outcome of a long unpriced relationship is not that the client leaves when you raise it; it is that resentment builds on your side until the account becomes one nobody wants to staff. That is a worse ending for both parties than a rate conversation.

If the increase is large because it has been deferred for years, split it and say so: "We're moving to £X now and £Y in twelve months, so it isn't a single jump." Predictability is worth more to most clients than the absolute number.

What to do this month

1. Calculate your real break-even hourly cost, using genuine available hours. Most agencies find they are 20-40% below where they should be.
2. Rank every client by margin, not revenue. Identify the bottom two.
3. Raise your new-client rate today. No notice needed, no risk, immediate learning.
4. Put a date in the calendar for the existing-client increase, at the next natural boundary, with 60 days' notice.
5. Write the four-sentence notice now, while you are thinking about it clearly, rather than in the moment when the temptation to soften it is strongest.

The increase itself takes an afternoon. Everything hard about it happens before you send it.