An MVP that takes twelve months defeats its own purpose: by the time you finish, the market has moved and the hypotheses have gone stale. Ninety days is enough — if you cut in the right places.
What fits into 90 days
One main scenario that creates value (an order, a booking, a calculation), sign-in, one key integration and an admin for managing data. Weeks 1–2: discovery and prototype. Weeks 3–11: development with a demo every Friday. Weeks 12–13: deployment, analytics, handover. Budget €42,000–68,000.
What does not belong in an MVP
Multiple languages if you sell in one market. A mobile app if the web is enough. Five permission roles when ten people will use the system for the first six months. Each of these is a legitimate phase two — in phase one it just postpones the day you learn the truth from users.
What happens after 90 days
Launch day is not the goal, it is the start of measurement. The first month shows what users actually do — and that data, not opinions in a meeting, defines phase two. Sometimes it confirms the direction, sometimes it reveals the most-wanted feature is something entirely different. That is the whole point of an MVP: deciding by behaviour, not by presentation.
How a 90-day rhythm changes the vendor
A fixed 90-day frame with a demo every Friday forces the vendor to prioritise — they cannot hide behind “still working on the architecture”. You, meanwhile, see progress every week and can correct course before the budget runs out. A vendor who resists this rhythm tells you more about themselves than about your project.
The hardest part of a 90-day MVP is not the code. It is the discipline to say “not this, not now” — and for that you need a vendor who can push back, not nod along.