How Long Does It Take to Build an MVP? Week by Week
Six to twelve weeks for a real product with a real team. Longer than that and either the scope is not an MVP or the team is not shipping. This is the week-by-week plan we run, including the two weeks where projects usually die.
- Published
- Updated
The eight-week plan
| Week | Phase | What ships | Your job |
|---|---|---|---|
| 1 | Discovery | Scope doc, user flows, architecture, fixed price | Decide fast; answer within 24h |
| 2 | Design | Clickable prototype of every core screen | Test it on 3 real users |
| 3–4 | Build I | Auth, data model, first two flows on staging | Use staging weekly, file bugs |
| 5–6 | Build II | Remaining flows, admin, integrations, emails | Prepare launch list & content |
| 7 | Hardening | QA pass, security review, load test, analytics | Recruit first 20 users |
| 8 | Launch | Production deploy, monitoring, handover docs | Ship it. Watch the dashboard |
Add two to four weeks for marketplaces and two to eight for AI-native products, where the evaluation loop is its own workstream. Anything past sixteen weeks should be re-scoped into two releases.
Where projects slip
Almost every late MVP we have rescued lost its time in the same three places.
- Week 1 decisions dragging into week 4. Every unanswered question in discovery costs three days downstream. A founder who answers in an hour ships a month earlier than one who answers in a week.
- Design revisited during build. Once engineering starts, design changes are 5× the cost. Test the prototype hard in week 2 and then freeze it.
- Integration surprises. The third-party API that "definitely supports that" does not. Spike every integration in week 3, not week 6.
How to ship faster
- One decision-maker on your side, with the authority to say no.
- A component library, not a custom design system, for v1.
- Two user roles maximum. Admin can be a database view.
- Deploy to staging from day one. Momentum is visible progress.
- Cut the last 20% of scope in week 5 if it threatens week 8. You can add it back in week 9.
We commit to a launch date in the week-1 proposal and price the build fixed against it. Describe your product in the form below and we will send that proposal — timeline, pod, price — within 48 hours.
Frequently asked questions
Can an MVP be built in two weeks?
A validator — landing page, waitlist, one manually-run flow — yes. A product with accounts and data, no, not by a team you would want to keep.
What if I already have designs?
That saves most of week 2, provided the designs are complete and tested. Incomplete designs cost more time than no designs, because engineering fills gaps by guessing.
How much of my time does it take?
Plan on 3–5 hours a week: one review call, testing on staging, and fast answers in the shared channel. Founders who go quiet for a fortnight add a fortnight.
