Minimum Viable Product
The concept was introduced by Frank Robinson (2001) and popularized by Eric Ries in The Lean Startup (2011) as part of the Build-Measure-Learn loop.
What an MVP Is Not
A common misconception is that an MVP is simply a product with few features. The key lies in its purpose: an MVP is for learning, not for delivering. An MVP that does not generate validated learning is not an MVP: it is an unfinished product.
Types of MVP
There are multiple strategies for building an MVP, depending on the type of hypothesis to be validated:
- Concierge MVP: Instead of automating the service, it is delivered manually. This allows for the validation of the value proposition before any technology is built.
- Wizard of Oz MVP: The system appears automated but is actually operated manually behind the scenes. The user is unaware that no real technology is involved. This concealment has limits: the experiment must not lead the user to make decisions they would not otherwise make if they knew how it worked; their data must be handled with the same care as in any interview; and it is advisable to explain how it works once the interaction concludes. If any part of the responses is genuinely generated by AI, then—as of August 2, 2026—Article 50 of the European AI Act requires that users be informed they are interacting with an AI.
- Paper prototype or low-fidelity MVP: Mockups or sketches that simulate the product to gather feedback on the interface and user flow without developing anything.
- Landing page MVP: A landing page that describes the product and measures interest (via clicks or sign-ups) before the product actually exists.
- Explainer video MVP: A video that describes how the product will function, used to gauge interest without building the product itself.
- Single-feature MVP: A version containing only one feature of the complete product, released to measure usage and perceived value.
MVP and Generative AI
Generative AI tools have drastically reduced the cost of building certain types of MVPs:
- Functional software prototypes can be built in hours rather than weeks with coding assistants.
- Content MVPs (newsletters, courses, reports) can be produced with AI support at a fraction of the previous cost.
- However, the reduced cost of building does not eliminate the need for validation with real users: an MVP built quickly with AI but launched without feedback is still a failed MVP if the hypotheses are not validated.
Common mistake
Building an MVP that is too big. The MVP is not the first version of the product; it is the smallest possible experiment to learn something specific. If the team takes more than a few weeks to launch its MVP, it has probably built too much. The right question to ask before building any feature is: "What specific hypothesis are we validating with this?"
Resources
🏦 Skill Agile InceptionState of the art and development guide · Skill Arena
🏦 Product discovery and customer validationState of the art and development guide · Skill Arena
📄 Product Owner v.2.0 [in Spanish]Free download · Scrum Manager
📄 Mínimo Producto Viable (MVP): qué es y para qué sirve [Minimum Viable Product (MVP): What It Is and What It Is For, in Spanish]Scrum Manager Blog · May 2023
References
- Ries, Eric. (2011). The Lean Startup. Crown Business.
See also
Want to advance in agility? You can search for course and exam calls or go at your own pace by becoming a member of the Agile Club. This membership includes exclusive resources, e-learning classrooms, and access to Skill Arena: a space to practice and measure your agile skills at your own pace.