Churn Intervention Specialist
Checks
.agents/product-marketing-context.mdfor user segment profiles. Checks.agents/growth-metrics-context.mdfor churn baselines. If growth-mcp connected, usespredict_churn_riskfor real predictions.
Churn Types in Fintech
| Type | Signal | Best Intervention |
|---|---|---|
| Transaction dormancy | No tx in 14-30 days | Nudge + small voucher |
| Balance zero-out | Withdrew all funds | Reason survey + winback offer |
| Feature abandonment | Stopped using key feature | Re-onboarding flow |
| Competitor switch | Known competitor activity | Price match or exclusive benefit |
| Seasonal churn | Post-holiday drop | Streak reward, loyalty tier |
Save Offer Framework
Tier 1: Light nudge (no cost)
- Push notification with personalized message
- "We miss you" email with feature highlight
- In-app banner with trending feature
Tier 2: Small incentive (low cost)
- Time-limited voucher (e.g., 10% off next transaction)
- Bonus points on next action
- Free transaction fee waiver
Tier 3: Strong intervention (medium cost)
- High-value cashback on next 3 transactions
- Exclusive tier upgrade for 30 days
- VIP customer support access
Winback Sequence (Recommended)
| Day | Channel | Message | Cost |
|---|---|---|---|
| D0 (dormancy detected) | Push | "You have X points expiring" | Free |
| D3 | Push | "Trending near you: [popular mechanic]" | Free |
| D7 | In-app + Email | "Here's 10% off your next [action]" | Low |
| D14 | SMS | "Special offer: [exclusive cashback]" | Medium |
| D21 | Push | "Your [feature] is waiting" | Free |
| D30 | Survey | "Help us improve" + reason dropdown | Free |
Metrics to Track
- Save offer redemption rate
- Reactivated user D7/D30 retention
- Cost per reactivated user
- Incremental transactions from saved users
- Winback sequence conversion by step
Related Skills
- retention-analyzer — diagnose churn root cause
- growth-mcp-connect — pull churn predictions
- campaign-brief — write reactivation campaign brief
- voucher-mechanic-designer — design winback mechanics