Building a Customer Health Score System That Works

Customer health scoring is one of the most useful tools in a SaaS company's retention arsenal. A well-designed health score tells you which customers are thriving, which need help, and which are at risk of churning, often weeks or months before they leave.

In this guide, we'll walk through how to build a customer health score system that actually works, what metrics to include, and how to use it to drive retention and expansion.

What Is a Customer Health Score?

A customer health score is a composite metric that predicts customer outcomes, typically retention, expansion, or churn. It combines multiple leading indicators into a single score (often 0 to 100) that gives your team an instant snapshot of each customer's relationship with your product.

Try our customer health score calculator to see how different factors combine into a health score.

Key Components of a Health Score

The most effective health scores combine four categories of data:

1. Product Engagement

How actively does the customer use your product? Key metrics include:

• Daily/weekly active users (DAU/WAU)
• Core feature adoption rates
• Time spent in product
• Key action completion (e.g., "sent first campaign," "created first report")

2. Customer Sentiment

How does the customer feel about your product and relationship? Sources include:

• Net Promoter Score (NPS): use our NPS calculator
• CSAT survey responses
• Support ticket feedback
• Executive sponsor sentiment (from QBRs)

3. Support Interactions

How the customer interacts with your support team can reveal a lot:

• Number of support tickets submitted
• Ticket severity and resolution time
• Frequency of escalations
• Satisfaction with support interactions

4. Account Health Signals

Broader business signals that indicate account stability:

• Billing history (on-time payments vs. delinquencies)
• Contract status (months until renewal)
• Stakeholder changes (champion departure is a major red flag)
• Usage trends (declining usage over 30 to 60 days is predictive of churn)

Building Your Health Score Model

Here's a step-by-step approach to building your first health score:

1. Identify your outcome metric: Define what a "healthy" customer looks like (e.g., renews at higher tier, or renews at all)
2. Gather historical data: Look at customers who churned vs. those who expanded. What signals differed?
3. Select leading indicators: Choose 3 to 7 metrics that are measurable, leading (not lagging), and actionable
4. Weight and normalize: Assign weights to each component based on predictive power. Normalize to a 0 to 100 scale
5. Define thresholds: Green (80 to 100), Yellow (50 to 79), Red (0 to 49), or your own categories
6. Validate and iterate: Test your model against historical outcomes and refine weights

Using Health Scores in Practice

Once your health score is running, integrate it into your daily operations:

• Customer Success workflows: Trigger automated outreach for accounts that drop from green to yellow
• Sales prioritization: Focus expansion efforts on high health score accounts with low penetration
• Product roadmap: Use health score data to identify features that correlate with high retention
• Executive reporting: Report portfolio health score distribution to the board monthly

Common Mistakes

Avoid these pitfalls when building your health score:

• Too many inputs: More than 7 to 10 metrics creates noise, not signal
• Using lagging indicators: Revenue data is backward looking; focus on engagement and sentiment
• Not segmenting: A health score that works for enterprise customers may not work for SMB
• No action framework: A health score without automated workflows is just a report

Remember: the best health score is the one your team actually uses. Start simple, prove value, and iterate.

Build your health score with our Customer Health Score Calculator.