Lean Startup & MVP

Origin

The approach was formulated by Eric Ries (The Lean Startup, Crown Business, 2011), who combined his experience as a founder with the principles of lean production. Conceptually it builds on Steve Blank's customer development (The Four Steps to the Epiphany, 2005), whose central injunction — “get out of the building” — demands testing assumptions with customers rather than at the desk. Ries added the vocabulary that has since prevailed: validated learning, innovation accounting, and the pivot as a methodically grounded change of course.

Typical use

The approach belongs wherever market data are missing because the market does not yet exist: new product categories, new customer segments, new business models — in startups and corporations alike. In practice it follows on from modelling work: the critical assumptions of a Business Model Canvas or Value Proposition Canvas supply the list of hypotheses that the experiment loop works through. For optimising a running business with reliable data, by contrast, it is the wrong tool.

Procedure

1. Make the assumptions explicit

   The business idea is decomposed into its leap-of-faith assumptions: which value hypothesis (customers want this) and which growth hypothesis (it reaches enough customers) must hold for the venture to work? The riskiest assumption is tested first.

2. Build an MVP

   For that assumption, construct the smallest experiment that measures real customer behaviour — a landing page, a concierge service, a manually operated prototype. The yardstick is not product maturity but learning gained per week and per euro.

3. Measure with innovation accounting

   What is measured is behaviour, not applause: cohort analyses, and for each experiment one metric with a threshold defined in advance. Vanity metrics — cumulative sign-ups, page views — are expressly excluded, because they rise under almost any circumstances.

4. Learn: pivot or persevere

   At a fixed cadence, the team decides on the basis of the data: stay the course or pivot — replace a core assumption such as customer segment, problem or revenue mechanism, without discarding what has been learnt. A pivot is a result of the method, not an admission of failure.

Limits and typical mistakes

The most widespread misunderstanding concerns the MVP: it is a measuring instrument, not a half-finished product. Whoever ships a stripped-down version of the planned product without it testing a defined assumption is launching prematurely, not learning in a validated way — and may damage the brand with the very first customers. The second risk is metric theatre: experiments are designed to confirm what has already been decided, or numbers are arranged into a success story after the fact. Third, the approach has limits of applicability — with long development cycles, heavy regulation or safety-critical products, the loop cannot be shortened at will. And finally, Lean Startup does not answer the question of where: which search fields are strategically relevant must be settled beforehand.

Relation to the Innovator's Dilemma