The Lean Startup

Ries defines a startup as a human institution designed to deliver a new product or service under conditions of extreme uncertainty, and argues that its central activity is learning what customers want. He proposes validated learning as the unit of progress, measured by experiments rather than by shipped features or business plan milestones.

The method runs on a build-measure-learn loop. A team states its leap-of-faith assumptions about value and growth, builds a minimum viable product to test them, measures with actionable rather than vanity metrics, and then decides to persevere or pivot. Ries defines several types of pivot, including zoom-in, customer segment, platform and business architecture pivots, and argues that the pivot decision should be scheduled rather than left to a crisis. He also introduces innovation accounting, which sets a baseline, tunes the engine, and then evaluates whether the improvements justify continuing.

The book carries a strong operational component drawn from lean manufacturing. Ries argues for small batch sizes, continuous deployment, the andon cord, and five whys root-cause analysis adapted to software teams. His engines of growth framework identifies sticky, viral and paid growth as three distinct models with different governing metrics.

The evidence is anecdotal, drawn mainly from IMVU, where Ries was a founder, plus consulting cases. The IMVU story has been questioned, and the minimum viable product concept has been widely misused to justify shipping poor products. Thiel and others argue that iteration cannot substitute for a strong initial hypothesis, and several large companies succeeded by doing the opposite of what the book recommends. The framework works best where customer demand is genuinely unknown and worst where the technical risk dominates.