Choose a return window for a growth experiment readout
Match the review window to how the intervention produces results.
Identify when the main effect should appear
A clearer checkout message can affect same-day orders. An onboarding change may take weeks to affect retention. State when an outcome is expected and how much time each cohort needs.
Keep short- and long-term questions separate
Use early measures to verify that delivery and initial behavior changed. Use mature measures for the decision that depends on refunds, retention, or contribution. Resist replacing a slow outcome with a convenient early proxy.
Plan how to handle incomplete data
Mark recent cohorts as provisional. Compare only equivalent ages and record a follow-up date. If an urgent rollout decision is needed, name the evidence that remains pending.
Try a related tool
Find the share of eligible visits that completed a defined action. Count at most one conversion per visit for this calculation.
