Most of what we build ships under someone else’s logo. Agencies come to us with a signed statement of work and a delivery date, and the client on the other end never learns our name. People ask whether that bothers us. It doesn’t — but it does change how you have to work.
The credit you give up buys something back
When your name is on the footer, a certain amount of the sale is your reputation. When it isn’t, every engagement has to stand on the work alone. That turns out to be a useful constraint. There is no portfolio to coast on and no case study to point at, so the only argument available is the state of the repository on the day you hand it over.
It also means the failure modes are different. An agency that resells your work absorbs your mistakes in front of their own client, which is a far more expensive place for a mistake to land than an invoice dispute.
The partner who has to explain your bug to their client will only do it once.
What changes in practice
Four things we do differently because the work is white-label:
- No signature in the code. No vendor comments, no attribution in the readme, no telemetry phoning home. The codebase belongs to whoever paid for it, on day one, not after a handover call.
- Documentation written for a stranger. The person who reads it is not the person who briefed us, and often not even the agency — it’s whoever inherits the project in two years.
- Boring dependencies. A clever library we love is a liability to a team that has never seen it. We spend our novelty budget on the parts of the product nobody else could build.
- Estimates with the uncertainty attached. Our partner has to quote their own client off our number, so a confident wrong estimate is worse than a wide honest one.
The handover is the product
An agency’s real question is never “can you build this”. It’s “what happens to me when you leave”. Everything above is an answer to that second question. If you want to know how we write the documents that survive the handover, that’s its own post.