# What is an MVP, really?

Source: https://www.plutonapps.com/resources/what-is-an-mvp

Published 2026-09-14, updated 2026-10-24, sources checked 2026-10-24. By Milad Zarghani (https://www.plutonapps.com/authors/milad-zarghani).

**Minimum viable product (MVP):** A minimum viable product (MVP) is the smallest complete version of a product that real users can get real value from without the founder beside them: one kind of user, one journey, finished properly. It is not a cheap version of everything, and not a demo, but a real version of one thing.

Also known as: MVP, First release.

The phrase has been stretched until it means nothing. Here is a definition you can make decisions with.

> **The test** If one real person can get real value from it without you standing next to them, it is an MVP. If not, it is a demo.

## Where does the term MVP come from?

Eric Ries popularised it in the Lean Startup method. He defines the minimum viable product as the version of a new product that lets a team collect the maximum amount of validated learning about customers with the least effort. The MVP starts the build-measure-learn loop.

Build something, watch how customers react, then decide whether to pivot or persevere. Ries also warns that, despite the name, an MVP is not about creating minimal products. The point is what you learn from real use, not how little you built.

## How is an MVP different from a prototype?

A prototype tests whether the idea makes sense. You, testers and investors use it, the data is throwaway, and an AI builder such as Lovable is ideal for it. An MVP tests whether people use and pay for it: real users, unsupervised, with real data that must be kept safe and engineering underneath.

|  | Prototype | MVP |
| --- | --- | --- |
| Purpose | Learn whether the idea makes sense | Learn whether people use and pay for it |
| Who uses it | You, testers, investors | Real users, unsupervised |
| Data | Throwaway | Real, and must be kept safe |
| Built with | An AI builder such as Lovable is ideal | Engineering underneath: security, tests, releases |

## Why do most MVPs get too big?

Because scope is added one reasonable request at a time. The discipline is not saying no to good ideas; it is deciding which single journey has to work before any of them are considered. Everything else waits for evidence from people using that journey.

## Does the minimum in MVP apply to quality?

The minimum in MVP applies to scope, not to quality. A small product can still leak data, lose payments or break on a Friday night. Keep the feature list short and the engineering underneath it real. Our page on [what production-ready means](https://www.plutonapps.com/resources/what-does-production-ready-mean) lists what that engineering covers.

## How do you scope an MVP designed in Lovable?

Name one kind of user and one journey, design only that journey in Lovable and cut every other screen, decide what you need to learn and how to measure it, put real foundations under that journey, and release to a small group of real users, changing scope only on evidence from them.

1. Name the one kind of user and the one journey that has to work end to end, for example signing up, booking a slot and paying.
2. Design that journey in Lovable and cut every screen it does not need. Settings pages, dashboards and admin tools can wait.
3. Decide what you need to learn from the first users, and how you will measure it, before you release.
4. Put real foundations under that one journey: access rules on the data, tests, backups and monitoring.
5. Release to a small group of real users and change scope only on evidence from them.

The common mistake is the reverse: a prototype that demos every feature, none finished, released to real people with real data. If your Lovable app is already live and behaving like a demo, a [vibe coding rescue](https://www.plutonapps.com/services/vibe-coding-rescue) is the route back; if you are choosing who builds the next version, see [choosing a tech partner for an MVP](https://www.plutonapps.com/resources/choosing-a-tech-partner-for-an-mvp), and for what ongoing engineering costs, [our pricing](https://www.plutonapps.com/pricing).

## Frequently asked questions

### Can an AI-built prototype be an MVP?

It can become one. Once real users rely on it unsupervised, it needs the same foundations as any live product: access rules, tests, backups and monitoring. The minimum applies to scope, not quality: keep the feature list short and the engineering underneath real.

### What does Eric Ries mean by a minimum viable product?

Ries made the term central to the Lean Startup method. He defines it as the version of a new product that lets a team collect the most validated learning about customers with the least effort, and stresses it is not about minimal products.

### How long should it take to build an MVP?

As long as it takes to finish one journey properly, and no longer. Ries describes spending two weeks building a feature nobody wanted, and says in hindsight that was far too long. Test demand cheaply first, then build the smallest complete version that real users can rely on.

### What comes after an MVP?

Learning. In the Lean Startup loop you build, measure how customers actually react, then decide whether to pivot or persevere. Each new feature should answer something the first users taught you. The foundations under the MVP, such as security, tests and backups, carry forward rather than being rebuilt.

## Read next

- [Plans and pricing](https://www.plutonapps.com/pricing)
- [Our work: from Lovable design to live product](https://www.plutonapps.com/work)

## Sources

- [The Lean Startup: Methodology (Eric Ries)](https://theleanstartup.com/principles)
- [Wikipedia: Minimum viable product](https://en.wikipedia.org/wiki/Minimum_viable_product)
- [Eric Ries, Lessons Learned: Venture Hacks interview, What is the minimum viable product? (March 23, 2009)](https://www.startuplessonslearned.com/2009/03/minimum-viable-product.html)
- [Eric Ries, Lessons Learned: Minimum Viable Product: a guide (August 3, 2009)](https://www.startuplessonslearned.com/2009/08/minimum-viable-product-guide.html)