Lean Startup

Building companies by testing assumptions with small experiments instead of big plans.

Also known as: lean startup, lean methodology, Eric Ries

Lean Startup (from Eric Ries’s book) treats a startup as an experiment machine: write down your leap-of-faith assumptions, test them with the smallest thing that yields evidence, measure honestly, and persevere or pivot based on what you learn. Business plans pretend you know; lean methods assume you don’t and price learning cheaply.

plan-driven:   6-month roadmap → build → launch → learn (expensive if wrong)
lean:          assumption → smallest test → evidence → next assumption (cheap if wrong)

Its unit is validated learning — knowledge about customers demonstrated by a change in behavior (a signup, a payment, a repeat visit), not by opinions collected. Everything else (code, decks, press) is waste until it produces that.

The classic mistakes:

Lean theatre. Calling any rushed hacking “lean” while testing nothing and measuring vanity metrics. Lean is disciplined experimentation, not speed for its own sake.
Testing what is easy instead of what is risky. Optimising button colors while the riskiest assumption — that anyone pays at all — sits untested. Riskiest first, always.
Pivoting without learning. Changing direction every month from restlessness rather than evidence is flailing, not leaning. A pivot follows a validated conclusion, documented.
Applying it where execution matters more. Once the model works, the job is scaling and quality — endless experimentation becomes dithering. Lean finds the model; discipline runs it.

Where it fits: pre-fit work — discovery, MVP, early iteration. Pair it with fast iteration speed: the method sets the questions, shipping speed sets how quickly you get answers.