Definition
A customer health score is a composite metric that combines multiple signals to predict whether a customer will renew, expand, or churn. It gives your team a single number that represents the overall state of each customer relationship.
Why it matters
No single metric tells the full story. A customer might log in every day but never use your core features. Another might have low usage but just signed a three-year contract. Health scores combine these signals into one view so your team can act before problems become cancellations.
During onboarding, health scores are especially valuable. Early signals like onboarding progress, support ticket volume, and feature adoption predict long-term retention with surprising accuracy. A customer who's struggling in week two won't magically recover in month six.
Teams that track health scores during onboarding can intervene early. They can reassign resources, adjust the onboarding plan, or schedule a check-in call before the customer disengages completely.
How to measure it
A health score aggregates multiple inputs into a single number (typically 0-100). The most common approach is a weighted average.
Common inputs
- Product usage: Login frequency, time in app, actions completed
- Onboarding progress: Steps completed, time to value, completion rate
- Support activity: Ticket volume, ticket severity, response satisfaction
- Engagement: Email open rates, webinar attendance, community activity
- Sentiment: NPS, CSAT, or customer effort score responses
- Financial signals: Payment history, expansion conversations, contract terms
Simple three-signal health score for onboarding
Here's a starter model you can build in a spreadsheet:
| Signal | Weight | Score range |
|---|---|---|
| Onboarding completion (%) | 40% | 0-100 |
| Product usage (weekly logins) | 35% | 0-100 |
| Support sentiment (CES) | 25% | 0-100 |
Formula:
Health score = (Onboarding completion x 0.40) + (Usage score x 0.35) + (Sentiment score x 0.25)
Example: A customer at 80% onboarding completion, 60 usage score, and 90 sentiment score would get: (80 x 0.40) + (60 x 0.35) + (90 x 0.25) = 32 + 21 + 22.5 = 75.5
Scoring methods
- Weighted average: Assign weights to each input. Simple to build and explain.
- Rules-based: Set thresholds (e.g., "if usage drops below X for two weeks, mark as at-risk"). Easy to customize.
- ML-based: Train a model on historical data. More accurate but harder to maintain.
Start with a weighted average. You can always add complexity later.
Interpreting the score
- 80-100 (Healthy): Customer is on track. Continue normal engagement.
- 50-79 (Needs attention): Something isn't right. Review the underlying signals and reach out.
- 0-49 (At risk): Immediate action needed. This customer is likely to churn without intervention.
How to use health scores during onboarding
Track health scores from day one. OnboardingHub gives you built-in progress analytics so you can see exactly where each customer stands in their onboarding flow. Pair that data with your CRM to build a health score that catches problems early.
Review scores weekly during the first 90 days. After that, monthly reviews are usually enough for healthy accounts.
Related terms
- Customer effort score: A survey metric that measures task difficulty. Often used as an input to health scores.
- Net revenue retention: The revenue you keep from existing customers. Health scores help predict which accounts will expand vs. churn.
- Customer churn rate: The percentage of customers who leave. Health scores are designed to predict and prevent churn.
Want to track the signals that predict customer success? Start with the customer onboarding metrics guide for a full breakdown of what to measure and when.