Retention debates collapse when “active” is undefined. Pick a behavior that maps to value — completed a task, synced data, or reached a paid unlock — and hold that definition for at least one quarter.
Acquisition-week cohorts beat calendar-month blends when you are diagnosing a launch. Month blends hide week-one onboarding bugs behind later traffic.
Always publish the exclusion rules: bots, internal accounts, and users who never passed the paywall. Leaving them in makes healthy products look broken and broken ones look fine.
When sample size is thin, say so. A wide confidence band is more honest than a sharp line that nobody should trust.