βΆLean vs Agile β aren't they the same?
Agile is a development methodology (iterative cycles, daily standups, sprints); Lean adds customer validation to each cycle. Agile β Lean. You can be agile (shipping fast) while building the wrong thing. Lean enforces: before each sprint, define a testable hypothesis; after sprint, measure real customer behavior; if hypothesis fails, pivot. Together: Agile (how to build) + Lean (what to build).
βΆWhat's an MVP and why does mine keep becoming a real product?
MVP = Minimum Viable Product that tests ONE hypothesis (not all features). Example: Dropbox MVP was a video of file syncing (not the app itself) β tested 'would users care?'. Common mistake: building an MVP that's just a smaller version of your vision. Fix: write your riskiest assumption first ('Users will pay for X'); MVP only tests that. If it fails, pivot; if it passes, plan the next MVP.
βΆHow many Build-Measure-Learn cycles should I run before deciding to pivot?
No fixed number. Metrics-driven: run until you have statistical confidence (usually 30-100 customers minimum). Innovation accounting: track validated learning per dollar spent; if learning per dollar stops improving after 3-5 cycles, pivot. Speed: some teams do 10 cycles/month (SaaS onboarding tests), others do 1-2/month (hardware). The point: pivot based on evidence, not gut feel, and iterate faster than competitors.
βΆWhat's the difference between a pivot and failure?
Pivot = you validated a core hypothesis but discovered a different customer need. Failure = you ran the experiment, hypothesis was wrong, and you have no new direction. Both are learning; pivot is worse-case learning with a path forward. Measure pivots: 'customer acquisition cost too high for startups, but enterprise loves it' = pivot to enterprise. 'Nobody clicks the feature' = probably fail, not pivot.
βΆWhy did Toyota Production System lead to Lean Startup?
TPS (1950s) introduced kaizen (continuous improvement), Just-In-Time (make only what's needed), and problem-solving culture. Eric Ries applied TPS principles to startups: kaizen = Build-Measure-Learn, JIT = MVP (make minimum viable), culture = validated learning > opinion. Lean Startup isn't new; it's TPS for software/services.
βΆWhat metrics should I measure in a Lean experiment?
Avoid vanity metrics (total signups, page views); they don't imply action. Actionable metrics: conversion rate, churn, retention, NPS, time-to-value, cost-of-acquisition. For hypothesis 'users will pay for X': measure willingness-to-pay (% of users signing up), not traffic. Track: baseline (before) β hypothesis β metric change β decision (pivot/persevere).
βΆWhen is Lean NOT the right approach?
Lean assumes uncertainty. Use it for new products, new markets, new customer segments. Don't use for: known problems (just build), operational efficiency (use Six Sigma), or scaling proven product (switch to Agile/waterfall). Example: launching a new feature in a mature product = Lean (is market need real?). Rolling out infrastructure for known feature = not Lean (execute, don't validate).