The Lean Startup method, introduced by Eric Ries, treats every startup decision as a hypothesis to be tested cheaply through Build-Measure-Learn loops. The core unit is the experiment: a small build, a measurable outcome, and a decision to persevere or pivot based on the data.
The practice popularized concepts like the MVP, validated learning, innovation accounting, and the explicit pivot. Most modern accelerators teach a watered-down version of it, but the underlying insight, that startup work is mostly hypothesis testing under uncertainty, not project execution, remains the dominant frame for early-stage company building.