Health score
helth skor
A composite metric that combines usage data, engagement signals, and support patterns to indicate whether a customer account is healthy or at risk.
A health score is a single number or color (green/yellow/red) that summarizes the state of a customer account. It combines multiple signals into one actionable indicator: is this account healthy, at risk, or in trouble?
Typical health score inputs: product usage (logins, feature adoption, API calls), support engagement (ticket volume, severity, CSAT), business engagement (QBR attendance, executive sponsor status), and financial signals (payment history, contract terms).
The danger of health scores is false precision. A number that combines fifteen inputs weighted by arbitrary coefficients looks scientific but may not reflect reality. Validate your health score against actual churn data. If accounts scored green churn at the same rate as accounts scored yellow, your score is broken. Iterate on the model quarterly.
Examples
A simple three-color health score.
Green: daily active users above 70% of licensed seats, support CSAT above 4.0, and executive sponsor engaged. Yellow: one or two metrics below threshold. Red: all three metrics below threshold. The CS team reviews all red accounts weekly.
A health score catches a declining account.
Account health drops from green (score 85) to yellow (score 62) over two months. Primary driver: daily active users dropped from 45 to 18 out of 50 licensed seats. The CSM calls and learns that the customer reorganized a team. 25 users moved to a different division. The CSM proposes a plan to onboard users in the new division.
A health score model gets recalibrated.
After six months, the CS team reviews the model. Finding: support ticket volume has zero correlation with churn. Customers who submit tickets are engaged, not unhappy. The team removes ticket volume and adds feature adoption rate. Model accuracy improves from 65% to 78%.
In practice
Read more on the blog
Frequently asked questions
What inputs should a health score include?
Start with three to five signals: daily active usage, feature adoption breadth, support CSAT, executive sponsor engagement, and payment history. Weight them based on which signals correlate most strongly with actual renewal outcomes. Keep it simple and validate quarterly.
How often should you review health scores?
The score should update automatically (daily or weekly). The CS team should review red accounts weekly and yellow accounts bi-weekly. The model itself should be recalibrated quarterly against actual churn data to ensure the weights and thresholds are correct.
Related terms
Using usage data, support patterns, and engagement signals to identify customers likely to cancel before they actually do.
The team and practice of ensuring customers achieve their goals with your product. Owns retention, expansion, and advocacy.
The rate at which customers cancel or do not renew. Measured as logo churn (customers lost) or revenue churn (dollars lost).
When an existing customer extends their contract for another term. The foundation of recurring revenue.
The percentage of revenue retained from existing customers after expansion, contraction, and churn. Above 100% means you grow without new sales.
Want the complete playbook?
Picks and Shovels is the definitive guide to developer marketing. Amazon #1 bestseller with practical strategies from 30 years of marketing to developers.