lean-startup

Start by identifying your riskiest leap-of-faith assumption, then pick the smallest MVP type that tests it: smoke test for demand, concierge for value, or single feature for engagement. Define the metric that validates or invalidates the hypothesis before building anything.

Vanity metrics like total signups or page views always trend up but drive no decisions. Actionable metrics show clear cause and effect, such as signup-to-active conversion rate or cohort retention, and directly inform what to change next.

Pivot when experiments consistently fail to validate hypotheses, metrics stay flat despite iterations, or progress is too slow for your runway. Persevere when metrics improve and learning accumulates. Define pivot triggers upfront before running experiments.

The sticky engine relies on high retention and low churn, the viral engine requires a viral coefficient above 1.0, and the paid engine needs LTV exceeding CAC by at least 3x. Focus on one engine at a time rather than combining them.

Lean Startup fits uncertain new products where demand is unproven. It is less suitable for well-understood problems with stable requirements, regulated builds with fixed specifications, or pure technical feasibility work, where prototyping or design sprints fit better.