The Lean Startup

Eric Ries argues that a startup is not a smaller version of a real company — it's an experiment under conditions of extreme uncertainty. The right unit of progress is not features shipped or revenue earned but validated learning: each cycle, did your hypotheses about customer, problem, and solution become more or less likely to be true? Build → Measure → Learn replaces plan-and-execute. The MVP isn't a smaller product; it's the smallest experiment that produces a falsifiable answer.

Lean Startup's gift to first-time founders is permission to fail faster. You don't need a finished product before you talk to customers. You don't need a perfect plan. You need a hypothesis you can falsify within two weeks. The discipline isn't about lowering quality — it's about lowering waste. Every feature you build before validation is potentially a tax on the runway you don't have.

The pivot-or-persevere frame is the second gift: every cycle ends with a real decision, not a vague 'keep going' default. Validated learning is the only scoreboard that matters until product-market fit. For Phase 1 founders specifically — still building toward a launch — Ries recalibrates the question from can we build this? to should we build this, and what's the smallest test that tells us?