Time-and-materials billing puts the risk on the client. A fixed price for the whole project puts it on the agency, which then protects itself with change requests. Neither produces a good conversation.
We quote one phase at a time. Discovery has a price. At the end of it, with a written brief in hand, we quote Design. At the end of Design, with real pages in the browser, we quote Build. Each quote is fixed. Each phase ends with a demo, and the client can stop after any of them and keep everything produced so far.
Why the risk stays with us
If a phase costs us more than we estimated, we absorb it. That has happened. It is our estimate that was wrong, not the client’s fault, and a client who is charged for our mistake will not come back.
What we do not absorb is a change in scope. If the site gains a second language halfway through Build, the next phase is re-quoted. The current one is not.
Why the client can leave
Because it keeps us honest. A studio that knows the client can stop after Discovery writes a Discovery worth paying for. A studio that has the whole budget signed on day one does not have to.
In nine years, six clients have stopped after Discovery. Four came back later with a bigger project. The brief we wrote them was good enough to take elsewhere, and that is the point.
The time it went wrong
In 2023 we quoted a Build phase at ten weeks for a catalogue of what the client called “about two hundred products”. It was two thousand, with a data model that changed weekly. We had not asked to see the spreadsheet.
We finished it in nineteen weeks and charged for ten. The project was profitable for the client and not for us, and the Discovery template now has a line that says: look at the data, not the description of the data.
What this costs the client
Slightly more than the cheapest quote they will receive, because we price the phase we can see, not the one we hope for. And in exchange, an invoice that matches the proposal, every time.
