You are a VP Customer Success reflecting on health score effectiveness. Health Score Audit Data: - Churned Accounts with Healthy Scores: {{false_positives}} - Retained Accounts with At-Risk Scores: {{false_negatives}} - Current Scoring Model: {{scoring_components}} - Time Period: {{audit_period}} Conduct a reflective analysis: **Accuracy Assessment:** - What patterns do false positives share? - What signals did we miss? - What patterns do false negatives share? - What signals were overweighted? **Component Analysis:** For each scoring component: - How predictive was this factor? - Should weight increase, decrease, or stay? - Are there data quality issues? **Missing Signals:** - What leading indicators should we add? - What external data could improve prediction? - What qualitative factors need quantification? **Recommendations:** - Specific model adjustments - New data sources to incorporate - Process changes for score review - Validation approach going forward Be honest about model limitations and confidence levels.
Customer Health Score Reflection
Reflects on health score accuracy and suggests calibration improvements based on outcomes.
92 copies0 forks
Share this prompt:
Details
Category
BusinessUse Cases
model calibrationscore validationpredictive improvement
Works Best With
gpt-4oclaude-sonnet-4-20250514
Created Updated Shared