Description
Cohort retention analysis: curve types, churn segmentation (power/core/casual), Kaplan-Meier survival times.
Aggregate metrics lie. A 95% retention rate sounds healthy until you see that month-6 retention is 40% and the 'average' is pulled up by week-1 users who never come back. Cohort analysis reveals the real story. 1. COHORT DEFINITION: Group users by signup week. Each cohort = everyone who joined that week. Track their activity (login, purchase, key action) in each subsequent week. This gives a retention matrix: rows = cohorts, columns = weeks since signup, cells = % active. 2. RETENTION CURVE TYPES: - FLAT: 60%+ retained at month 12. Sticky product (Slack, Notion). Healthy. - DECAY to plateau: drops quickly (weeks 1-4) then stabilizes (20-30% at month 6). Typical SaaS. Target: plateau >25%. - LINEAR DECAY: drops every week with no plateau. User don't discover lasting value. Product fix needed, not marketing. - TOOTHPICK: 100% week 1, <5% week 2. Viral one-off gimmick. Must acquire endlessly. 3. CHURN SEGMENTATION (break the plateau into segments): - POWER USERS (top 10% by activity): churn <1%/mo. These are your reference customers. - CORE (next 30%): 3-5%/mo churn. Target for re-engagement. - CASUAL (bottom 60%): 15-20%/mo churn. Either not a fit, or missing onboarding. 4. SURVIVAL ANALYSIS (Kaplan-Meier): For each segment, compute median lifetime. Power users: median 18 months. Core: median 8 months. Casual: median 2 months. If the gap between core and casual is >6 months, invest in onboarding. If power user and core are close, invest in advanced features. 5. LEADING INDICATOR: Which week's retention predicts month-6 retention? Usually week-2. If week-2 retention <40%, month-6 will not recover regardless of re-engagement. OUTPUT: Retention matrix heatmap, curve classification, 3-segment churn table with survival times, leading indicator.
No comments yet. Be the first!