GLOSSARY · PRODUCT EXPERIMENTS SOURCE-REVIEWED

Minimum viable product (MVP)

The smallest credible product or experiment that lets a team learn whether a key customer and business assumption holds.

PLAIN-LANGUAGE DEFINITION

An MVP is an early product form used to begin learning about a customer problem and a proposed solution. “Minimum” describes the smallest scope that can produce meaningful evidence; “viable” means a participant can complete a relevant task. It need not be a polished commercial product, and a prototype or manual service may sometimes test an assumption better.

The smallest credible product or experiment that lets a team learn whether a key customer and business assumption holds.

What the term means

An MVP is an early product form used to begin learning about a customer problem and a proposed solution. “Minimum” describes the smallest scope that can produce meaningful evidence; “viable” means a participant can complete a relevant task. It need not be a polished commercial product, and a prototype or manual service may sometimes test an assumption better.

Why it matters to venture decisions

The Lean Startup describes a build-measure-learn loop: identify a problem, develop an MVP to begin learning, measure what happens and decide what to change. The point is not simply to ship quickly. It is to make a test that can change a product or venture decision.

A practical way to use the concept

Write the assumption first: who has the problem, what task they need to complete, what outcome matters and why the proposed approach might improve it. Select the lowest-risk experiment that can observe that outcome. Define the participant, baseline, measure, duration, data boundary and decision rule in advance. A concierge workflow, clickable prototype, synthetic-data demo or narrow technical benchmark may each be appropriate for different questions.

Limits and cautions

An MVP is not a shortcut around safety, security, accessibility, privacy or regulatory review. A test with friendly users is not evidence of repeat use or scalable economics. For AI products, define representative cases, human review, failure handling and a baseline before testing.

Sources and review

General educational information only; not legal, financial, investment or professional advice. Reviewed 31 October 2026. Recheck the source for updates before relying on it in an operational decision.