A focused MVP takes 3 to 6 weeks if the scope is one user, one job, and one payment path. It takes a quarter if you are building a platform. The difference is almost never the framework. It is whether you cut before you start.
The week-by-week
Week 1. Foundations. Auth, data model, billing skeleton, and a deploy pipeline on a real URL. If this week slips, cut from week 3. Do not extend.
Week 2. The activate moment. The one workflow that delivers value, end to end, with real data. Ugly is fine. Working is not optional.
Week 3. Money and polish. Payments live, one onboarding email, error states, analytics events, launch checklist.
Weeks 4–6. Only if you have a second user type you cannot fake, an integration the activate moment depends on, or an AI pipeline that needs evaluation, not just a prompt.
What makes it take months
- The word “platform” in the one-job statement
- Two user types that both need to log in on day one
- An integration required to show any value
- Designing the pricing page before the product exists
- A cut list with fewer than five items on it
How we keep it inside a month
We run Discovery in week one as a worksheet, not a workshop series. The MVP Scoping Playbook is that worksheet: one-job statement, loop map, cut list, calendar.
If a feature can be done by a human for the first 50 customers, it is not in the MVP. Settings pages, role matrices, admin dashboards, and “AI insights” almost always fail that test.
What “done” means
A URL a stranger can use. A card they can be charged on. Events that tell you whether they came back. A repo you own. If any of those are missing, you have a demo.
If you want a date against your idea, bring the one-job statement to a call. We will tell you whether this is a three-week build or a six-week one, and what we would cut to keep the earlier date.