mkt-analyticslisted

# Marketing Analytics — Measurement Design **Decide what you measure, when, and how first; pick the tool afterward.** Tools (Google Analytics, PostHog, etc.) are just options — without a plan, any tool produces garbage. ## KPI Tree (top → bottom) 1. **North Star** — a direct proxy for the value the product creates (e.g. weekly active teams, not logins) 2. **Input metrics** — the 3~5 behaviors that move the North Star (early activation rate, weekly retention, paid conversion rate) 3. **Diagnostic** — used to trace root causes. Anything below level 3 stays off the dashboard (50 metrics = 0 metrics) ## Event Design Rules - Event naming: `object_action` in lowercase snake_case (`checkout_started`); no past tense, no duplicates - Cap core events at **≤15**. Pull context from properties instead of adding events - Every event must carry either `user_id` (post-login) or an anonymous ID (pre-login) — otherwise sessions can't be stitched - Explicitly separate micro- vs macro-conversions: only macro (purchase etc.) counts as a KPI; micro (clicks etc.) is diagnostic ## Quality Audit Checklist - No double-fired events (triggered from both modal and redirect paths) - Test traffic and internal IPs filtered out - No conflicts between per-channel UTM auto-rules and manual tagging - Debugging: verify one real user's journey captures end-to-end on the timeline before shipping ## Output ``` # Measurement Plan — {product} ## KPI Tree (North Star → inputs → diagnostics) ## Event spec tabl