Most early-stage product builds either ship late or ship wrong. The 90-day MVP framework we use at PVP is designed to avoid both outcomes. Here's how it breaks down.
Weeks 1–2: Define what done looks like
Almost every bad MVP comes from skipping this step. Before any code is written, we agree in writing on three things: the single user journey we're shipping, the customer who will validate it, and the measurable signal that tells us it worked.
If we can't agree on those three things in two weeks, the project doesn't proceed. It's not the right time to build.
Weeks 3–6: Architecture and core build
Most teams over-engineer at this stage. We deliberately under-engineer: smallest schema that supports the user journey, cheapest hosting that meets the SLA, no admin tooling beyond what the team needs to operate. Anything more is wasted effort because the MVP's job is to be killed and replaced.
By the end of week 6, the core user journey works end-to-end on a real (not seeded) database. Nothing else is built.
The MVP's job is to be killed and replaced.
— Shahan Ali Naqvi, Chief Technology Officer
Weeks 7–9: Real users, fast iteration
This is where most founder energy gets misallocated. The temptation is to add features. The actual job is to put the MVP in front of five to ten paying or paying-intent customers and watch what they do.
We run a weekly user session, write up what we learned, and ship the highest-priority change before the next session. Three cycles of this is usually enough to tell us whether we're on to something or whether we should pivot.
- 5–10 paying or paying-intent users
- Weekly customer sessions
- Highest-priority change shipped each week
- Decision point: continue / pivot / kill
Weeks 10–12: Handover and decision
By week 10 we know enough to write a real strategy doc — what we learned, what to build next, what to drop, what budget is needed for the next 90 days. We hand the founder a fully documented codebase, the analytics dashboard, the customer call recordings, and a written next-quarter plan.
From that point the founder can either continue with PVP, take it in-house, or move to a different supplier. We've engineered the handover deliberately to make any of those three viable.
