Scope creep and the hours nobody quoted for

This is less a set of instructions about scope creep than an argument, and it is worth saying so at the start.

How it accumulates

Scope rarely expands in a single visible step; it grows through requests that individually cost twenty minutes and collectively cost a week. Each one arrives with an apology and a reason, which makes refusing feel disproportionate to what is being asked. By the time the total is obvious, a great deal of the extra work has already been delivered without charge.

The client is frequently unaware of the cumulative effect, because they experience the requests one at a time exactly as you do. That is why the problem is usually solved by a process rather than by a confrontation.

The signals that it has started

Requests that begin with a phrase minimising the effort involved are the most reliable early indicator. New people joining the client's side mid-project almost always bring opinions that were not in the brief.

For most people, a deadline that keeps moving while the deliverables keep growing is the standard shape of a job going wrong. Approval passing to someone who never agreed the original scope means the specification is effectively being rewritten. Noticing any of these early is worth more than any clause, because early conversations are far cheaper than late ones.

Pricing the change when it is requested

The workable response is neither refusal nor silent acceptance but a short note confirming the request and what it adds in time or cost. Framing it as information rather than resistance keeps the tone right, since the client is being told a consequence rather than being denied.

Put simply, many requests are withdrawn at this point, which demonstrates that they were never worth what they would have cost you. Where the client accepts the cost, the work becomes chargeable rather than absorbed, which is the entire objective. Doing this from the first small request sets the pattern, whereas starting at the tenth reads as a change of attitude.

Measuring the damage on fixed-price work

On a fixed fee there is no invoice line that reveals the overrun, so the only visibility comes from recording hours anyway. A simple log comparing estimated hours against actual hours per project turns a vague sense of unfairness into a number. That number is what allows you to price the next similar project correctly rather than repeating the mistake.

It also identifies which clients consistently overrun, which is frequently a stable characteristic rather than bad luck.

Without the log, a business can run for years at an effective rate far below what its price list implies.

The client relationships that cause it

Clients without a clear internal decision-maker generate the most creep, because each stakeholder adds requirements independently. Clients who bought on price often have expectations set by a more expensive supplier, which is a mismatch that surfaces mid-project. Long relationships drift, as informal favours accumulate into an assumed level of service that was never priced.

A periodic reset conversation, framed around what the current arrangement includes, is easier than it sounds and usually welcomed. Where a client will not accept any boundary, the honest calculation is what the account earns per hour against everything else you could do.

None of this is a substitute for talking to a clinician if something feels wrong.

Building the process into how you work

Confirm the scope in writing at the start, in language a non-specialist would recognise as a list. Include a defined number of revisions and state what happens after them, which converts an argument into an agreed rule. Send a short written note for every change, however small, and keep them in one place.

At the end, review estimated against actual and adjust the next quote, since the pricing lesson is the valuable output of a difficult job. None of this requires being difficult; it requires being consistent, which is considerably easier.

The takeaway

Price every change the moment it is asked for, and log hours on fixed-price work so the next quote is right.

Pick the one that costs you least, and let the rest wait.