funnel-analysislisted

# Funnel Analysis ## When to use this skill Use when measuring **sequential step-by-step conversion** through a defined flow. Triggers: - "Build a signup funnel" - "Where are users dropping off?" - "Conversion from step A to step B" - "Analyze the checkout flow" - "Activation funnel" For lifecycle behavior over weeks/months → `cohort-analysis`. For experiments → `ab-test-analysis`. ## Required inputs | Input | Why it matters | |---|---| | Funnel steps (ordered) | The sequence to measure | | Unit of analysis | User, session, or visit | | Time window | Conversion deadline between steps | | Strict vs non-strict order | Must steps occur in order, or just all eventually? | | Cohort filter | Which users to include (e.g., new signups only) | ## Workflow 1. **Define the steps explicitly.** Each step is an event name + filter conditions. Avoid vague steps like "engaged." 2. **Decide strict vs non-strict ordering.** - **Strict (sequential)**: step N must follow step N-1 in time. Standard for onboarding/checkout flows. - **Non-strict (any order)**: user must have done all steps eventually. Use for feature adoption funnels where order doesn't matter. 3. **Set the conversion window.** Between step N and N+1, what's the max time to convert? Common defaults: - Same session: ~30 min - Same day: 24h - Same week: 7 days - Lifetime: open-ended (but inflates conversion artificially) 4. **Compute the funnel using `scripts/build_funnel.sql`**, which produces: - Step