MVP: Beyond the Minimum Viable
MVP stands for Minimum Viable Product. In practice, many teams focus on the “minimum” and forget the “viable.” The result: products so basic that nobody wants to use them — and the MVP fails to validate anything.
What a Real MVP Is
An MVP validates a specific hypothesis with the minimum effort.
Hypothesis: “Users want to schedule services online instead of calling.” MVP: A landing page with a scheduling form + WhatsApp to confirm.
This already validates the hypothesis without building a full scheduling system.
Types of MVP
1. Landing Page MVP
A page describing the product with a “sign up” button. Measures interest before building.
2. Concierge MVP
A service performed manually for each customer. “The app does this” — in reality, there is a person behind it.
3. Wizard of Oz MVP
The user sees an interface, but behind the scenes everything is manual. Zappos started this way: the founder went to a shoe store, took photos, and put them online. When someone bought, he purchased the shoes at the physical store and shipped them.
4. Prototype MVP
A functional prototype (Figma, Webflow) to test usability before coding.
5. Single Feature MVP
A real product, but with a single feature. Twitter started as “what are you doing?” — that was it.
MVP by Sprints
Ideal flow for validation with agile sprints:
| Sprint | What to Deliver | Goal |
|---|---|---|
| Sprint 0 | Research + Story Map | Understand the problem and define hypotheses |
| Sprint 1 | Landing page + analytics | Validate interest |
| Sprint 2 | Core feature (1 functionality) | Validate the solution |
| Sprint 3 | Iteration based on feedback | Improve or pivot |
| Sprint 4 | Expand or discard | Decide with data |
Estimating an MVP
MVPs are estimated differently:
Use T-Shirt Sizing, Not Points
For MVPs, the precision of story points is unnecessary. “Is this feature M or L?” is enough.
Prioritize Learning Over Perfection
If an 8-point feature delivers the same learning as a 3-point one (simplified version), choose the 3-pointer.
Timebox Everything
Set a limit: “Let’s validate in 3 sprints. If there is no positive signal, we pivot.”
Common MVP Traps
The “MVP” That Is a Full Product
“Our MVP has login, registration, profile, dashboard, reports, integrations…” That is not an MVP, it is a product afraid to launch.
An MVP That Validates Nothing
“We launched the prototype” but with no success metrics defined. What is success? 100 sign-ups? 1,000 visits? Define it before launching.
An MVP Without an Iteration Plan
MVP launched, feedback received, and now what? The backlog of next steps needs to be ready.
MVP Metrics
- Conversion rate — how many visitors became users?
- D7 Retention — did they come back after 7 days?
- NPS — would they recommend it to someone?
- Engagement — do they actually use it or just clicked once?
- Qualitative feedback — interviews with 5–10 users
Agile MVP = Build-Measure-Learn
The lean cycle:
Build → Measure → Learn → (repeat or pivot)Each MVP iteration should last 1–3 sprints. More than that, and the cost of being wrong is too high.
Conclusion
An MVP is not a half-baked product — it is the cheapest way to learn. The faster you validate hypotheses, the less time (and points) you spend building something nobody wants. Estimate the MVP, but with a focus on learning, not perfection.