How to charge for out-of-scope work without losing the client

The question is not whether to charge. It is what you are charging for.

Ignition surveyed 273 agency managers and executives for The 2025 Agency Pricing & Cash Flow Report. The headline finding is the one everybody quotes, and it deserves to be quoted: 78% of agencies say they rarely or only sometimes charge for out-of-scope work.

It is tempting to read that as a failure of nerve, and almost every article about it does. But nerve is not the scarce resource here. Most agency owners are perfectly willing to be paid. What stops them at the keyboard is narrower: they do not know what number to write. "Some extra work happened" has no price, and a price requires having decided what unit you sell. A lot of agencies never decided.

So this post is not about confidence. It is about the mechanics — how to arrive at a number you can defend, which vehicle to put it in, what sentence to attach to it, and what to do in the specific moment when the client says no. The relationship survives all of this far more often than people expect, and the reason it survives is almost entirely about form.

Price the change, not the hours it cost you

The instinct is to count the hours the extra work consumed and multiply by a rate. It is the most natural method and the weakest one, because it invites the client into a conversation about your internal efficiency — whether six hours was really necessary, whether a faster designer would have taken four. You do not want that conversation. Here are three methods that avoid it, roughly in order of how well they hold up.

Price it as a unit you already sell

The strongest method, and available only if you did the work upfront. If your proposal said "two rounds of revision per page template", a third round has an obvious price: whatever a round costs, which you can state because you priced rounds when you priced the job. The client is not being asked to accept a new pricing concept mid-project. They are buying one more of something they already bought several of.

Price it as a proportion of the baseline

When the extra work is not a unit you sell, size it against the whole. "The original scope was five page templates for €9,000. This is a sixth, so €1,800." The arithmetic is visible, the logic is the client's own budget rather than your cost base, and it is hard to argue the sixth should be free when the first five had a price. It also protects your margin, because it inherits the margin in the original quote.

Re-quote the whole thing

For a genuinely large change — the platform moved, the audience changed, a second business unit appeared — do not bolt a change order onto a dead estimate. Say plainly that this is a different project from the one you quoted, and offer to re-quote it. It feels drastic and is usually the least confrontational option, because it stops you both haggling over increments of a job you privately agree no longer exists.

Four vehicles, and how to pick one

Having a price is not the same as having a way to charge it. The vehicle you choose does most of the work of keeping the relationship intact, because it decides how formal the moment feels. Four are worth knowing, and the mistake is using the heaviest one for everything.

• A change order. A named, priced, dated addition to the agreed scope, approved before the work starts. The right default for anything with a real price, and the only one of the four that leaves a paper trail if the relationship later goes wrong.
• A line on the next invoice, flagged in advance. For small, frequent, obviously-extra items where formal approval each time would be absurd. In advance is load-bearing: an unannounced line is the most reliable way to turn a €400 disagreement into a lost account.
• A trade. On a monthly retainer, the honest move is often a swap rather than more money: "The extra landing page fits this month if the blog post moves to next month — or I quote it separately. Your call." It makes a fixed budget visible without making it a fight.
• Absorbing it, out loud. Sometimes the right answer is no charge. Then say so: "That's about three hours outside what we scoped — we'll cover it this time." Announced free work buys goodwill. Silent free work buys nothing, and teaches the client the scope was never real.

The selection rule is simpler than it looks. If the work has not started yet and the price is meaningful, it is a change order. If it is small and the pattern is recurring, it is a flagged invoice line. If the budget is genuinely fixed, it is a trade. If you have decided to eat it, it is still a conversation — just a pleasant one.

What to actually write

The message that works is short, and it contains four things in this order. Anything longer reads as a justification, and a justification invites a negotiation.

1. Name the change as a fact, not a complaint. "This is a third round on the homepage; we scoped two." Not "we've been going back and forth a lot." One is checkable; the other is an emotion the client may dispute.
2. Give the price plainly, with no cushioning. "€480." Not "it would be around €480 or so, if that's okay?" A hedged price reads as an opening bid and gets treated as one.
3. State the schedule consequence. "It moves delivery to the 14th." The part people leave out, and often what actually persuades: it reframes the ask from money to trade-offs, and it is a fact about your calendar rather than an opinion about value.
4. Ask a closed question. "Want me to book it?" That can be answered with one word on a phone. "Let me know your thoughts" cannot — and that is how change orders go quiet for three weeks and become unbillable.

Assembled, that is four sentences: "That's a third round on the homepage — we scoped two. It's €480, and it moves delivery to the 14th. Want me to book it?" Ninety seconds to write. The length matters more than the wording, because a change order that takes half an hour to prepare will lose, every single time, to the thought that you will sort it out at the end of the project.

Two words to delete before sending, every time: just and sorry. "I'm just going to have to charge a little for this, sorry" tells the client you find the request unreasonable, which invites them to argue about whether it is. You are not apologising for having a price. You are quoting.

Money is easiest to discuss before the work, and it is not close

Agencies already know this instinctively about new business — almost nobody starts a project with no deposit and a vague sense that it will be settled later. Ignition's respondents bear that out: 49% request partial payment in advance, and 16% require payment in full upfront.

The odd thing is how completely that discipline evaporates for changes. The same agency that will not begin a €20,000 website without a deposit will happily do €2,000 of unagreed extra work across six weeks and raise it afterwards, if at all. There is no principled reason for the difference. The start of a project has a ritual, with a document and a signature. A Tuesday afternoon does not.

This is the mechanism behind every piece of advice above. Before the work, a price is a quote and the client is deciding whether to buy. After the work, the identical price is a bill for something already received, and their only remaining move is to dispute it. Nothing about the fairness of the number changed — only which side of the work the sentence landed on. Speed is not a nice-to-have here. It is the process.

When they say no

They will, sometimes, and a no is not a failure of the method. It is the method working: you asked a question and got an answer, which is strictly better than the silence you get from never asking. But "no" means three quite different things, and responding to the wrong one is what damages relationships.

"We don't have budget for that"

Usually true, and usually not about you. The correct response is a trade, not a discount: offer to swap the extra work for something already in scope, or to schedule it into next quarter's budget. Discounting here teaches an expensive lesson — that your prices move under mild pressure — and it does not even solve the client's actual problem, which is a budget line rather than a number.

"I thought that was included"

This one is genuinely ambiguous and worth slowing down for. Sometimes it is a negotiating position. Very often it is true: they did think it was included, because the proposal was vague and they read it the generous way. Go and look at what you actually wrote before replying. If the document is unclear, absorb it, say plainly that the scope was not specific enough, and fix the wording next time. You have bought a lot of trust for the price of one revision round.

"No — and I'm annoyed you asked"

Rare, and worth reading carefully rather than reacting to. Sometimes the relationship has a problem the change order merely surfaced. But sometimes it is the fourth such message this month, and the real problem is that you are sending too many small change orders — the client experiences your project as a meter running. Bundle them, raise fewer and larger ones, and check whether the original scope was too thin.

In every one of the three cases, you then do the thing that makes the next one easier: write down that the change was requested, priced, and declined. Not to be passive-aggressive about it — because in four weeks, when delivery is late, the question of why will have an answer that is a record rather than two people's memories.

Absorbing the work is a legitimate decision. Absorbing it silently is not.

Nothing here argues that every hour of drift should be billed. Some should not be: the estimate you got wrong, the client you want to keep for three more years, the favour that costs an hour and returns a referral. Judgement is the job. The problem is not generosity — it is generosity nobody records, which cannot be counted, cannot be pointed at later, and does not read as generosity to the person receiving it.

There is also a second cost, standing behind the first. The same Ignition survey found that 56% of agencies say they typically wait anywhere from two weeks to two months after the due date to be paid, and that 71% see at least one invoice in four paid late.

Make it routine, because routine is what actually holds

PMI's Pulse of the Profession 2018, a survey of 5,402 project professionals, ran a statistical analysis to find what most affects project success. Three things came out on top, and the second was avoiding scope creep or uncontrolled changes to a project's scope — ranked alongside executive sponsorship, not as a footnote to it.

Two things are worth taking from that pair of numbers. The first is that maturity genuinely moves the figure — eighteen points is a large effect for a management practice. The second is that even high-maturity organisations sit at 45%. Nobody controls this to zero. The realistic goal is not a project that never drifts; it is an agency where drift routinely produces a priced decision instead of a quiet donation.

1. Write the baseline in countable units — rounds, templates, pages, stakeholders, reviews. You cannot price a change against a baseline nobody can count.
2. Price the common extras in the proposal, not when you need them. A rate for an additional revision round, sitting in the original document, turns the hardest conversation in this post into an administrative one.
3. Give the change one owner. On a small team, whoever spots drift is usually whoever is doing the work — the worst-placed person to raise money. Decide in advance who writes the sentence.
4. Keep the vehicle light enough for a bad week. The test is not whether your process is rigorous, but whether it still happens on the Thursday when two projects are late and someone is ill.

None of this requires software, and for a small studio running one project at a time, a shared document and a habit are genuinely the right answer. What software changes is the last item on that list: whether the noticing survives the bad week. That is what we built Tave around — a baseline held as countable deliverables, watchers comparing what is happening against what you sold, and a finding that becomes a priced change order the client approves in their portal, available afterwards as an invoice line.

But the habit is the part that matters, and it is available this week without buying anything. Pick the three most common extras in your work. Write a price next to each. Put those prices in your next proposal. You have just removed the reason most change orders never get sent — which was never courage, and was always the blank space where a number should have been.