The discount nobody names: what scope creep costs a studio

A note on the evidence, because there is less of it than you would expect

There is no published dataset measuring how often architecture practices absorb additional work unbilled, or how many hours it costs them. We looked properly: the AIA, Deltek, PSMJ, Zweig and RIBA all publish benchmarking, and none of them measure it. Trade articles about scope creep in this profession routinely quote percentages with no source behind them at all.

What is actually measured

• Fifty two per cent of projects experienced scope creep, up from 43 per cent five years earlier. That is the Project Management Institute’s Pulse of the Profession 2018, drawn from 4,455 practitioners across industries, not architecture alone.
• Seventy five per cent of projects finished on or under budget in FY2025, which means one in four did not. Deltek Clarity, 896 architecture and engineering firms.
• Thirty three per cent of firms named a well defined scope among their top three project management strengths. Eighty per cent named client relationships. Same study.

That last pairing is the most revealing thing in the data. The profession is confident about relationships and unconfident about scope, and those two facts are causally linked. The better the relationship, the harder it is to answer a friendly request with a fee proposal.

The arithmetic, on your own project

Take a 50,000 fee at a twenty per cent margin, so 40,000 of cost and 10,000 of profit. Now absorb some additional work at a blended cost of 90 an hour.

Why architects absorb it

Not because they do not notice. Because the alternative costs something visible and the absorption costs something invisible. Raising a variation risks an awkward conversation today. Absorbing it costs margin that will not be measured until the year end, if at all.

There is also a scoping problem underneath. Additional services are only additional if the original service was written down precisely enough to sit beside them. Where the brief was loose, everything is arguably included, and the client is not being unreasonable when they assume so.

What to do instead

1. Write the exclusions, not just the inclusions. A scope that lists what is not included, how many design options are covered and how many revision rounds, converts an argument into a reference.
2. Price the variation at the moment it is requested, not at the end. A fee quoted before the work is done is a normal commercial conversation. The same fee raised afterwards feels like a penalty and gets negotiated.
3. Use a rate, not a favour. Publishing an hourly rate for additional services makes the next request a purchase decision for the client rather than a test of the relationship.
4. Track hours against the fee weekly, not at completion. The hour count where the project turns negative is knowable in advance. Knowing it at the end is just accounting.
5. Count what you absorbed anyway. Even if you do not charge, record it. A studio that can say it gave away 180 unbilled hours last year prices differently the following year, and the fee discount impact calculator shows what absorbing that volume did to the margin.