retention-analyzer

name	retention-analyzer
description	Analyze cohort retention, diagnose retention drops, identify at-risk segments, and recommend intervention strategies. Use when the user wants to understand retention trends, analyze cohort data, diagnose a retention drop, or plan retention campaigns. Specialized for fintech and super apps.
metadata	{"author":"thaolst","version":"1.0","license":"MIT","related_skills":["growth-mcp-connect","churn-intervention","campaign-brief","campaign-planning"]}

Retention Analyzer

Checks .agents/product-marketing-context.md for product context. Checks .agents/growth-metrics-context.md for baseline retention. If growth-mcp connected, pulls real cohort data via analyze_retention and predict_churn_risk. Otherwise, use expert defaults below.

Fintech Retention Benchmarks (SEA)

Metric	Good	Average	Poor
D1 Retention	> 50%	30-50%	< 30%
D7 Retention	> 30%	15-30%	< 15%
D30 Retention	> 20%	10-20%	< 10%
Monthly Churn	< 15%	15-30%	> 30%

Diagnostic Framework

When retention drops, check:

1. Seasonal effect — holiday spending spree → natural D1 dip
2. Campaign hangover — big promo → users wait for next promo
3. Feature regression — bug, UX change, performance issue
4. Competitor activity — competitor launched similar mechanic
5. Segment shift — acquired wrong user segment (incentive-driven)

Intervention Matrix

Problem	Intervention	Expected Lift
D1 drop (activation)	Onboarding flow fix, welcome voucher	+5-15%
D7 drop (habit)	Push nudge series, streak reward	+3-10%
D30 drop (churn risk)	Re-engagement campaign, winback voucher	+2-8%
General decay	Loyalty program, points economy	+5-20% over 3 months

Related Skills

• growth-mcp-connect — pull real cohort data
• churn-intervention — design save offers
• campaign-brief — write retention campaign brief
• voucher-mechanic-designer — design winback voucher