Agency scope creep is a pricing problem, and the ninety-day check that catches it

The account everyone likes is the one to worry about. The client pays on time, the team enjoys the work, and over eighteen months the monthly deliverable has gone from eight ad creatives to fourteen, plus a report, plus a Slack channel that gets answered at 9pm. Nobody decided any of that. Each piece was a reasonable yes, and the fee has not moved since the contract was signed.

That is what scope creep looks like in practice, and it is why most of the advice about it misses. The advice is about contracts. The problem is about margin.

Scope creep is a pricing problem that shows up in delivery

Every agency treats scope creep as a project management failure, which is why the standard fixes are a tighter statement of work, a revision cap, and a change-order form. Those are worth having. They also do very little, because the accommodations that cause the damage never look like a change request. They look like being helpful, and nobody files a change order for that.

What actually moves is the account's gross margin. Fee stays flat, hours rise, and the margin walks down from a healthy 55% to the low 30s without a single conversation. By the time it feels wrong, the client has a year of precedent that says the expanded scope is the scope.

The number that matters is the fully-burdened cost of every hour on the account, divided into the fee. If you have never run that per account, the retainer pricing method walks through it, and the fully-burdened cost entry explains why salary alone understates it by 25% to 40%.

What it is costing you

Ignition's 2025 Agency and Cash Flow Report found that 78% of agencies rarely or only sometimes charge for scope creep, and that 57% lose between $1,000 and $5,000 a month to out-of-scope work, with a further 30% losing more than that.

Take the low end of that range seriously. An agency losing $3,000 a month to unbilled work is losing a junior hire a year, and because the loss is spread across six accounts in small pieces, it never appears as a line anyone can point at. The margin floor is the only instrument that makes it visible.

The ninety-day check

Once a quarter, for every account, take last quarter's fees, subtract the fully-burdened cost of every hour that touched the account, and divide by the fees. Include account management, QA, and the owner's own time, which is the hour most agencies forget to count and the one most often spent on creep.

Sort the list by margin. Anything under 40% gets a conversation this quarter. Anything between 40% and 50% gets watched, because 50% to 60% is where a healthy account sits and an account drifting below 50% is usually an account that has already absorbed a year of unbilled scope.

Quarterly is the right interval because it is short enough to catch creep before it becomes precedent and long enough that the hours data means something. A monthly check reacts to one bad week. A yearly check finds out about the problem after the client has stopped believing the original scope was ever real.

Re-scope before you re-price

For an account under the floor, the order of operations is fixed, and skipping a step costs revenue.

Re-scope at the same fee. List what the account currently gets, mark what the fee was written to cover, and offer to return to that. This works more often than owners expect, because the client values the relationship and the core deliverable far more than the extras, and most of the extras were things they asked for once and kept receiving.

Re-price with ninety days' notice. If the client wants the expanded scope, the expanded scope has a price. Tie the reason to scope rather than to your costs, because your costs are your problem and the scope is a shared fact. Ninety days is long enough that the increase reads as fair and short enough that you are not funding the gap for another two quarters.

Release. Only after the first two have been declined. A client who wants the larger scope at the smaller fee has told you what they think the work is worth, and the decision on when to let a client go is now a margin fact rather than a relationship judgment.

The clause that helps and the clause that does not

A revision cap and a defined turnaround window belong in every retainer. They set the shape of the work and give the account manager something to point at when a request lands outside it.

A clause that says "additional work billed at $X per hour" does almost nothing, because the person who would have to invoke it is the same person who said yes to the request in the first place. The contract can describe the scope. Only the margin check can tell you the scope has moved.

The accounts to check first

Run the check hardest on the accounts with the best relationships. That sounds backwards, and it is the point: creep concentrates where the accommodations are made willingly, by people who like the client and never escalate a favor to anyone tracking margin. The difficult account that argues about every deliverable is usually the one with the cleanest scope, because every addition has already been negotiated.

If you want this run across your own accounts before you decide anything, the diagnostic does exactly that.

What it looks like when it works

Fraggell Productions grew from £23k to £110k a month over 24 months. One of the early moves was a pricing calculator built off the existing price sheet so that every package could be tested for margin before it went out, and a pod compensation model that held gross margin above 70% on every pod. The average monthly retainer went from £5.5k to £13.5k over the same period, because once the scope of each package was priced correctly, the expanded work the clients wanted had a number attached to it instead of being absorbed.

Read next

• When to fire a client, and how to know before it costs you a year
• How to price an agency retainer so it survives contact with delivery