How is a Churn Health Score calculated?

The Churn Health Score starts at 100. For each churn insight the AI surfaces from your cancellation feedback, points are deducted based on severity: 20 points for critical insights, 12 for high, 6 for medium, and 2 for low. The score floors at 0. A product with two critical and two high insights would score 100 − 40 − 24 = 36.

What is a good Churn Health Score for a SaaS company?

A score of 80 or above (grade A) indicates strong retention health with only minor churn drivers. A score of 65–79 (grade B) is solid — real issues exist but none are structural. Below 50 (grades C, D, F) signals that churn problems are serious enough to address before scaling acquisition spend.

How is a Churn Health Score different from a churn rate?

A churn rate measures the volume of customers or revenue lost over a period. A Churn Health Score measures the severity and nature of the underlying causes. Your churn rate is the outcome; the Churn Health Score is the diagnostic. A company can have a 3% monthly churn rate with a score of 78 (minor, fixable issues) or a score of 42 (deep structural problems not yet reflected in financials at scale).

Who invented the Churn Health Score?

RetentionCheck invented the Churn Health Score as the headline output of its AI-powered churn analysis product. The scoring methodology — severity-weighted deductions from a 100-point baseline — was designed to give SaaS founders a single, actionable number rather than a flat list of churn themes.

What Is a Churn Health Score?

What a Churn Health Score Is

A Churn Health Score converts AI-analyzed cancellation feedback into a single 0–100 number. The number tells you, at a glance, how severe your current churn drivers are. A score of 82 means your product has some issues but nothing structurally broken. A score of 31 means you have at least one critical retention problem that will compound if you don't act on it this sprint.

RetentionCheck introduced the Churn Health Score as the headline output of every churn analysis. Before it existed, founders got a list of themes. A list of themes doesn't tell you whether you have a minor UX friction or a pricing model that's structurally misaligned with your market. The score does.

How the Score Is Calculated

The calculation starts at 100 and deducts points based on the severity of each churn insight the AI surfaces from your cancellation feedback.

Insight Severity	Points Deducted	What It Means
Critical	−20	A fundamental product, pricing, or positioning problem driving significant cancellations
High	−12	A clear, recurring friction point affecting a material portion of churned customers
Medium	−6	A real issue mentioned by multiple customers but not dominant in exit feedback
Low	−2	An edge case or minor complaint from a small subset of churned customers

The score floors at 0. A product with five critical insights doesn't score −100; it scores 0 and earns an F grade. The floor keeps the number interpretable as a health percentage rather than a punishment metric.

A typical analysis surfaces 5–8 insights. A product with two critical insights and two high insights would score 100 − 40 − 24 = 36, a D. A product with one high and three medium insights would score 100 − 12 − 18 = 70, a B.

Grade Thresholds

The score maps to letter grades to make the number immediately actionable without requiring a mental model of what 67 vs. 71 means.

Grade	Score Range	Interpretation
A	80–100	Strong retention health. Churn drivers are minor or low-severity. Focus on growth.
B	65–79	Solid foundation with fixable friction points. One sprint of targeted improvements can move this to an A.
C	50–64	Meaningful retention problems exist. At least one high-severity driver needs immediate owner and deadline.
D	35–49	Serious retention risk. Multiple high or critical issues. Churn will compound without structural changes.
F	0–34	Retention emergency. Critical issues are dominating exit feedback. Stop acquiring until you fix the leaks.

Why Severity-Weighted Scoring Matters

A flat count of churn themes is misleading. Five low-severity complaints about your documentation are not the same as one critical complaint about pricing misalignment. Treating them equally in a ranked list makes the documentation issue look as urgent as the pricing problem if it appears more often.

Severity weighting forces the score to reflect business impact. A critical insight — say, customers consistently reporting that your core feature doesn't do what they expected when they signed up — carries 10x the weight of a low insight about a missing keyboard shortcut. That asymmetry is intentional and accurate.

See how to analyze cancellation feedback for a breakdown of what makes an insight critical vs. low severity.

How RetentionCheck Generates the Score

When you paste cancellation feedback into RetentionCheck, the AI reads every response and clusters the underlying reasons customers left. For each cluster, it assigns a severity level based on three signals: how often the theme appears, how directly it links to the cancellation decision, and how fundamental the issue is to the product's core value proposition.

The severity assignments are deterministic given the prompt constraints — the same feedback analyzed twice will produce the same or near-identical severity labels. The Churn Health Score is then computed as a pure mathematical function of those severity labels. There's no curve, no normalization against other companies, and no randomness in the final number.

You can see example analyses with scores to calibrate what a B vs. a D looks like on real feedback datasets.

What Score to Target

A score above 80 (grade A) means your churn drivers are manageable. Focus on acquisition and expansion rather than retention surgery. A score between 65 and 79 (grade B) is where most growing SaaS products live — there are real issues, but none are existential.

A score below 50 (grade C or worse) is a signal to pause growth spending. You're filling a leaky bucket. Acquiring customers at CAC only to lose them in 90 days because of an unaddressed critical issue is one of the most expensive mistakes a SaaS company can make in years two through four.

For context on what churn rates correspond to these score ranges, see what is a good churn rate for SaaS and churn rates by industry.

Using the Score Over Time

A single score is a snapshot. The score becomes a retention instrument when you track it across analysis runs. If you analyze your feedback monthly, you should see the score move when you ship fixes for the issues the AI identified. A score that stays flat or declines after shipping fixes means either the fixes didn't land with customers or new churn drivers have emerged.

This is why RetentionCheck stores your score history and surfaces trend data for Pro users — so the score functions as a leading indicator of retention health, not just a post-mortem metric. See cohort retention analysis for how to layer score trends with cohort data to isolate where problems originate.

Churn Health Score vs. Other Retention Metrics

The Churn Health Score is a diagnostic metric, not a financial one. It tells you about the quality and severity of your churn drivers, not the volume of customers lost or revenue impact. It complements, rather than replaces, metrics like net revenue retention, monthly churn rate, and customer lifetime value.

Think of it this way: your churn rate tells you the outcome. The Churn Health Score tells you the prognosis — whether the causes of that outcome are being addressed or are getting worse. A company can have a 3% monthly churn rate and a score of 78 (minor issues, addressable) or a score of 42 (deep structural problems that haven't yet hit the financials at scale). Those two situations require completely different responses.

For a full picture of retention health, track the Churn Health Score alongside net revenue retention and customer lifetime value.