Building a full-featured product before learning whether customers want it is risky. A minimum viable product, or MVP, is the smallest version that allows a team to collect the maximum amount of validated learning with the least effort. It contains only the core features needed to test the critical assumptions about value and growth.

The MVP is not a mediocre product; it is a deliberate experiment. Early users provide feedback that guides the next iteration. Features that do not contribute to learning are deferred. The approach is central to lean-startup methodology and is widely used in software, though the principle applies more broadly.

Defining what is truly minimum requires discipline. Teams often overestimate how many features are essential. A well-designed MVP answers a specific question: will customers use and pay for this core solution? The answer determines whether to invest further, pivot, or stop.