The churn radar for B2B SaaS·Book a call·Setup in 10 minutes·Trusted by CS teams·SOC 2 · GDPR · AES-256·
The churn radar for B2B SaaS·Book a call·Setup in 10 minutes·Trusted by CS teams·SOC 2 · GDPR · AES-256·
Fundamentals

Customer Health Score Formula: Build Your Own

Learn the math behind health scores—from basic averages to weighted formulas that actually predict churn.

12 min readUpdated February 2026

What is a Health Score Formula?

A customer health score formula is the calculation that transforms raw customer data into a single health score (typically 0-100). It defines which metrics you track, how you normalize them, and how they combine into the final score.

Think of it as a recipe: the ingredients are your customer metrics (usage, engagement, support), and the formula is how you combine them. A good formula accurately predicts which customers will renew and which will churn.

Why the formula matters: A poorly designed formula gives false positives (flagging healthy customers as at-risk) or false negatives (missing customers who are about to churn). The right formula gives you accurate, actionable predictions you can trust.

1. Input Metrics

Raw data points: login frequency, feature usage, support tickets, etc.

2. Normalization

Convert each metric to a consistent 0-100 scale for combination.

3. Weights (Optional)

Multipliers that make important metrics count more than others.

4. Aggregation

How you combine metrics into a final score (sum, average, weighted).

=The Basic Health Score Formula

The simplest health score formula is an unweighted average of normalized metrics:

Health Score = (M₁ + M₂ + M₃ + ... + Mₙ) / n × 100

Where M = normalized metric (0-1) and n = number of metrics

Step-by-Step Calculation

1

Collect raw metric values

Example: Customer logged in 15 times this month

2

Normalize to 0-1 scale

If 20 logins = perfect (1.0), then 15 logins = 15/20 = 0.75

3

Repeat for all metrics

Logins: 0.75, Feature usage: 0.80, Support tickets: 0.60

4

Average and scale

(0.75 + 0.80 + 0.60) / 3 × 100 = 71.7 health score

Normalizing Different Metric Types

Metric TypeNormalizationExample
Higher = Bettervalue / max_valueLogins: 15/20 = 0.75
Lower = Better1 - (value / max_value)Tickets: 1 - (2/5) = 0.60
BooleanYes = 1, No = 0Training complete: 1.0
PercentageAlready normalizedFeature adoption: 0.80

Weighted vs Unweighted Scores

The basic formula treats all metrics equally. But in reality, some metrics matter more than others for predicting churn.

Unweighted

All metrics contribute equally to the score.

Score = (M₁ + M₂ + M₃) / 3

Best for: Getting started, when you don't have churn data yet

Weighted

Important metrics count more than others.

Score = W₁×M₁ + W₂×M₂ + W₃×M₃

Best for: Mature teams with churn correlation data

The Weighted Formula

Health Score = (W₁×M₁ + W₂×M₂ + ... + Wₙ×Mₙ) × 100

Where W = weight (sum = 1) and M = normalized metric (0-1)

Example weighted calculation:

  • • Product usage: 0.80 × 0.35 weight = 0.28
  • • Feature adoption: 0.70 × 0.25 weight = 0.175
  • • Support health: 0.90 × 0.20 weight = 0.18
  • • Engagement: 0.60 × 0.20 weight = 0.12

Health Score = (0.28 + 0.175 + 0.18 + 0.12) × 100 = 75.5

How to Determine Weights

1. Analyze Churned Customers

Which metrics were consistently low before churn? Those get higher weights.

2. Calculate Correlation

Use statistical correlation to measure how strongly each metric predicts churn.

3. Start with Benchmarks

Usage 35%, Adoption 25%, Support 20%, Engagement 20%. Refine over time.

Automated alternative: Modern tools like FirstDistro use pre-tuned weighted formulas based on industry best practices. You get accurate health scores from day one without manual tuning—plus AI-generated recommendations for what to do about at-risk accounts.

Choosing Your Metrics

Include 4-7 metrics that are: measurable, actionable, and predictive. Here's a recommended breakdown:

Product Usage (30-40%)

Login frequency, time in-app, core action completion rate

Feature Adoption (20-25%)

% of key features used, breadth of usage, advanced features

Support Health (15-20%)

Open tickets (inverse), days since last ticket, severity trend

Engagement (15-20%)

Email engagement, community participation, training completion

Metrics to Avoid

  • Vanity metrics: Total logins ever (doesn't reflect current health)
  • Unactionable metrics: Company size (you can't change it)
  • Lagging indicators: NPS alone (by the time it drops, damage is done)

Formula Examples by Business Type

B2B SaaS (PLG / Self-Serve)

Score = (0.40×usage) + (0.25×adoption) + (0.20×recency) + (0.15×support)

Focus on product signals since you have limited human touchpoints.

B2B SaaS (Sales-Led / Enterprise)

Score = (0.30×usage) + (0.20×adoption) + (0.20×relationship) + (0.15×support) + (0.15×commercial)

Include relationship and commercial signals alongside product data.

Common Formula Mistakes

1

Too Many Metrics

Including 15+ metrics creates noise. Stick to 4-7 high-signal metrics.

2

Never Recalibrating

Review quarterly. A formula from 12 months ago may miss new signals.

3

One-Size-Fits-All

Enterprise and SMB customers have different healthy behaviors. Segment your formulas.

4

No Validation

Always check: do high-score customers actually renew? If not, something's wrong.

Tools That Calculate Automatically

Building and maintaining a formula takes significant effort. These tools handle the calculation automatically:

ToolApproachPricing
FirstDistroAI-powered, auto-learns weightsFrom $99/mo
GainsightRule-based, manual weights$50k+/year
ChurnZeroRule-based with templatesQuote-based
VitallyRule-based + ML optionsPer-user pricing

Why AI-Powered Beats Rule-Based

  • Auto-learn weights based on which signals predict churn for your customers
  • Continuously improve as more customer data comes in
  • Detect patterns humans might miss (specific feature combinations)
  • Eliminate bias from intuition-based weight assignments

Frequently Asked Questions

What is the basic customer health score formula?

The basic formula is: Health Score = (Sum of Metric Scores) / (Number of Metrics) × 100. Each metric is normalized to a 0-100 scale. For example, if you track login frequency (70), feature usage (80), and support tickets (60), your health score would be (70+80+60)/3 = 70.

Should I use weighted or unweighted health scores?

Start with unweighted scores for simplicity, then move to weighted once you have enough data to know which metrics best predict churn. Weighted scores are more accurate but require calibration. Most mature CS teams use weighted formulas with 4-6 metrics.

How many metrics should I include in my health score formula?

Include 4-7 metrics. Fewer than 4 misses important signals; more than 7 creates noise and makes the score harder to act on. Focus on metrics you can actually influence: usage frequency, feature adoption, engagement, and support interactions.

How do I weight metrics in my health score formula?

Weight metrics based on their correlation with churn. Analyze which metrics churned customers had in common. Typically: Product usage 30-40%, Feature adoption 20-25%, Engagement 15-20%, Support 10-15%, Commercial signals 10-15%. Adjust based on your data.

How often should I recalculate health scores?

Real-time or daily calculation is ideal for catching problems early. Weekly or monthly updates often miss critical warning signs. Modern tools like FirstDistro calculate continuously as customer activity happens.

What's a good health score threshold?

Standard thresholds are: Healthy (70-100), At-Risk (50-70), Critical (0-50). However, you should calibrate based on your actual churn data. If customers with scores above 60 rarely churn, your 'healthy' threshold might start there.

Can I have different formulas for different customer segments?

Yes, and you should. Enterprise customers have different healthy behavior patterns than SMB customers. A startup might log in daily while an enterprise team logs in weekly—both can be healthy. Create segment-specific formulas or adjust thresholds by segment.

Skip the spreadsheets. Get automated health scores.

FirstDistro calculates health scores automatically using proven weighted formulas. No manual tuning required—plus AI-generated recommendations for at-risk accounts.

Setup in 30 minutes • HubSpot & Attio integration • Cancel anytime

Related Guides