Why we ship in nine days, not nine weeks.
A first build inside two weeks isn't a stunt. It's the only way we've found to compress the feedback loop until the product can defend itself. Here's the operating model — and what it costs.
The first time a client sees real software is the moment the conversation stops being about ideas and starts being about decisions. Everything before that — the briefs, the slides, the discovery calls — is a tax on the work. We pay it as little as possible. By day nine, the build is in their hands and the questions are good ones.
This isn’t agency theater. Nine days is short because we removed the things that make projects long, not because we work faster than anyone else. Most of what passes for project management in our industry is choreography for people who haven’t agreed on the problem yet.
The cost of the long ramp.
A typical agency engagement opens with two to six weeks of discovery, branding, and wireframes. Everyone signs off on the documents. Then the build starts and roughly half of what got signed off turns out to be wrong — sometimes obviously, more often subtly. The team writes change requests. The timeline slips.
This is not a failure of discipline. It’s a failure of feedback. No artifact short of running software tells you whether the product solves the problem. Wireframes lie. Prototypes lie a little less. Real builds, exposed to real users, are the only honest answer.
If the first thing you ship is a deck, you are not a studio. You are a consultancy with a Figma seat. — Studio operating principle, rev. 3
What “nine days” actually means.
Day zero is the first call. Day nine is a working build deployed to a real environment that the client can use, click through, and break. It is not a demo. It boots, it persists state, it talks to whatever it needs to talk to.
1. Scope is fixed by week one.
The first 48 hours are spent narrowing. We don’t list features — we name the one thing the product has to be true about by day nine. Everything else is parked.
2. The stack is boring.
We use the same five or six tools, and we have the boilerplate ready. Reaching for something novel adds days, and days are what we don’t have. npx create-procesino@latest is a real command.
3. There is no project manager.
The person writing the code is the same person on the call. There is no translation layer between what the client asks and what gets built.
What it costs.
Nine days isn’t free. There are real trade-offs and we tell clients about them on the first call:
- The build is not finished on day nine. It’s defensible. It will run, it will get feedback, and it will be wrong in instructive ways.
- You can’t change the scope after day two. Or you can, but the new scope replaces the old one. We don’t add.
- The team is small. Usually one builder, sometimes two. There is no bench to escalate to.
If you have a thing that needs shipping — a small product, an internal tool, a playable ad, a stalled rebuild — and you’d rather skip the long ramp, tell us about it. The first call is free and short.