Back to Glossary
Metrics & Analytics

Customer Health Score

A composite metric predicting how likely a customer is to renew, expand, or churn.

Definition

Customer health score is a calculated metric that predicts customer outcomes based on multiple signals. It typically combines product usage data, support interactions, billing status, and engagement patterns into a single number or grade. A high score indicates a happy, engaged customer likely to retain. A low score signals churn risk requiring intervention.

Why It Matters

You cannot personally monitor every customer, but you can monitor health scores. Health scores help customer success teams prioritize who needs attention. They also power automated interventions. When a health score drops, you can automatically trigger emails designed to re-engage the customer before they churn.

How It Works

Define the signals that predict retention and churn for your business. Weight them based on predictive power. Common signals include login frequency, feature usage, support tickets, NPS responses, and billing issues. Calculate a score for each customer. Set thresholds that trigger alerts or automated actions.

Best Practices

  • 1Include both leading indicators (usage declining) and lagging ones (missed payment)
  • 2Weight signals based on actual correlation with outcomes
  • 3Recalibrate your model as you gather more data
  • 4Use health scores to trigger automated email sequences
  • 5Create different intervention playbooks for different score ranges
  • 6Track whether your score actually predicts churn
  • 7Share health scores with customer success and sales teams

Engagement Tracking

Track subscriber engagement and product usage to build your own health signals. Use these to trigger re-engagement campaigns automatically.

Learn More

Frequently Asked Questions

Common inputs include product usage metrics (logins, feature usage, time in app), engagement signals (email opens, support tickets), billing status (payment issues, contract details), and sentiment data (NPS, CSAT). Weight them based on what actually predicts churn in your data.

Backtest against historical data. Check if customers who churned had declining health scores before cancellation. Track false positives (healthy customers who churn) and false negatives (low scores who retain). A good model catches most churn risk with few false alarms.

Trigger re-engagement sequences when scores drop below thresholds. Start gentle (checking in, offering help) and escalate if the score continues falling. For rising scores, consider upgrade or referral asks. The goal is proactive intervention before churn becomes inevitable.