Product
Retain

AI Churn Prediction

Use PostHog data to build machine learning model predicting churn risk for proactive intervention.

CaptureProductEmail
$npx gtm-skills add product/retain/churn-prediction-model

Outcome

Predict top 20% at risk

Leading Indicators

  • Prediction accuracy
  • Churn rate
  • Intervention success

Instructions

  1. Define success metric and pass threshold (Predict top 20% at risk).

  2. In PostHog, create events for key actions; ensure app tracks all interactions.

  3. Instrument critical path with clear PostHog event names.

  4. Build minimum viable version with simple UI.

  5. Launch to 10-20 users via PostHog feature flag.

  6. Track progression in PostHog; create funnels for drop-off analysis.

  7. After 7 days, compute metrics vs. threshold.

  8. Review PostHog data for patterns and friction.

  9. If passed, document and proceed to Baseline; if not, iterate.

  10. Record observations for optimization.

Recommendations

Time

5 hours over 1 week

Play-specific cost

Free

Tools

PostHogProduct Analytics