health-scoring-effectiveness.md
Health Scoring Effectiveness
On-demand reference for the expansion-revenue-architect skill.
Does health scoring actually work? The data says yes — but only when done well.
| Finding | Data Point | Source |
|---|---|---|
| Health score accuracy (best platforms) | 8.9/10 rated accuracy | ChurnZero 2025 |
| Segment-specific scores | 15-20% more accurate at predicting at-risk accounts | Totango 2024 |
| CS platform impact on churn | 15-25% churn reduction within first year of deployment | Industry composite 2026 |
| Health score calibration | 50% of churned customers had "red" score 90 days prior (good); 80%+ (best-in-class) | CS Ops Reference |
What predicts churn vs. expansion:
- Gainsight finding: Satisfaction (relationship quality) is as predictive as usage
- Totango finding: Product adoption (feature depth + frequency) drives the strongest churn prediction
- Combined: Multi-signal health scores (usage + engagement + support + satisfaction) outperform single-dimension scores
CSM Coverage Model (from Gainsight Pulse 2025)
| Touch Model | Accounts per CSM | ARR per CSM | Source |
|---|---|---|---|
| High-touch | ~22 accounts | $2-5M ARR | Gainsight 2025 |
| Mid-touch | ~49 accounts | $2-5M ARR | Gainsight 2025 |
| Low-touch | ~144 accounts | Varies | Gainsight 2025 |