Metrics to track for a cloud-based B2B SaaS customer health score must combine product engagement signals (login frequency, feature adoption breadth), relationship signals (executive sponsor responsiveness, QBR attendance), and financial signals (invoice payment timing, expansion conversations) into a single composite score that predicts renewal probability 90 days in advance.
A customer health score that only measures product engagement misses the accounts most likely to churn: the ones where the champion has left the company, the executive sponsor has deprioritized the tool, or the billing contact has started routing invoices to a new cost center review. Product engagement is necessary but not sufficient for renewal prediction.
The Three-Layer Customer Health Score
H3: Layer 1 — Product Engagement Signals (40% weight)
Login frequency:
- Green: Active users log in ≥3x per week
- Yellow: Active users log in 1-2x per week
- Red: Active users log in <1x per week or MAU has declined >30% month-over-month
Feature adoption breadth:
- Green: Account uses ≥4 core features
- Yellow: Account uses 2-3 core features
- Red: Account uses 1 core feature (single-feature dependency is a churn risk)
Seat utilization:
- Green: >70% of purchased seats active in last 30 days
- Yellow: 40-70% of seats active
- Red: <40% of seats active (unused licenses are the most reliable churn predictor)
Core workflow completion rate:
- Measure the % of accounts completing the primary workflow weekly
- Red: <50% completion rate (product is not embedded in their workflow)
H3: Layer 2 — Relationship Signals (35% weight)
Executive sponsor engagement:
- Green: Executive sponsor attended last QBR and responds to outreach within 48 hours
- Yellow: Executive sponsor is present but engagement is declining
- Red: Executive sponsor has changed, departed, or stopped responding
Champion activity:
- Green: Internal champion actively advocates for the product (invites colleagues, submits feature requests)
- Red: Champion has left the company or changed roles to a non-user function
Support ticket sentiment:
- Green: Support interactions are feature requests and enhancement questions
- Yellow: Support interactions are bug reports and how-to questions
- Red: Support interactions are complaints, escalations, or expressions of intent to evaluate alternatives
NPS score trend:
- Green: NPS ≥30 and stable or improving
- Yellow: NPS 0-30 or declining
- Red: NPS <0 or account has submitted a detractor response in the last survey cycle
H3: Layer 3 — Financial Signals (25% weight)
Invoice payment behavior:
- Green: Invoices paid on time consistently
- Yellow: Invoices occasionally late (30-60 days)
- Red: Invoice overdue >60 days or invoice dispute initiated
Expansion conversation status:
- Green: Active expansion conversation in progress
- Yellow: No expansion conversation but account is in growth phase
- Red: Account has explicitly declined expansion or reduced seats at last renewal
Contract renewal timeline:
- Yellow at 120 days pre-renewal: Begin health assessment and renewal conversation
- Red at 90 days pre-renewal: Account health is Yellow or Red — trigger CSM intervention
Customer Health Score Composite Formula
Health Score = (Product Engagement Score × 0.40) + (Relationship Score × 0.35) + (Financial Score × 0.25)
Score range: 0-100
Green: 70-100
Yellow: 40-69
Red: 0-39
H3: Health Score to Action Mapping
| Score | Status | CSM Action | |-------|--------|------------| | 70-100 | Healthy | Quarterly check-in + expansion conversation | | 50-69 | At risk | Monthly check-in + feature adoption intervention | | 30-49 | High risk | Bi-weekly check-in + executive sponsor outreach | | 0-29 | Critical | Weekly executive escalation + save plan activation |
FAQ
Q: What is the most predictive metric for B2B SaaS customer health scores? A: Seat utilization — specifically the percentage of purchased licenses that are actively used in the last 30 days. Accounts with seat utilization below 40% are 3-4x more likely to churn at renewal than accounts above 70%.
Q: How often should you update a customer health score? A: Core product engagement metrics should update daily or weekly. Relationship signals should update after every touchpoint. Financial signals should update at invoice events. The composite score should be recalculated weekly.
Q: What is the difference between a customer health score and NPS? A: NPS is a single point-in-time survey signal capturing customer sentiment. A customer health score is a composite leading indicator calculated continuously from product usage, relationship signals, and financial behavior — it predicts renewal risk rather than measuring satisfaction.
Q: How do you weight the components of a customer health score? A: Start with equal weights (33/33/33) and calibrate against your actual churn data. Track which signals in churned accounts went red first and up-weight those signals. The 40/35/25 split in this guide is a starting point, not a universal truth.
Q: What is the earliest warning signal for B2B SaaS customer churn? A: Champion departure — when the internal user who drove initial adoption changes roles or leaves the company. This is often detectable 60-90 days before the churn decision, giving CSMs a window to re-establish the relationship with the new decision-maker.
HowTo: Build a Customer Health Score for a Cloud B2B SaaS Product
- Define the product engagement signals to track: login frequency, feature adoption breadth, seat utilization, and core workflow completion rate with Green/Yellow/Red thresholds for each
- Define the relationship signals: executive sponsor engagement, champion activity, support ticket sentiment, and NPS trend with threshold criteria
- Define the financial signals: invoice payment behavior, expansion conversation status, and days-to-renewal trigger points
- Build the composite score formula with initial weights (40% product, 35% relationship, 25% financial) and plan to calibrate against actual churn data after 6 months
- Map each score range to a specific CSM action: healthy accounts get quarterly check-ins, at-risk accounts get monthly intervention, critical accounts get weekly executive escalation
- Validate the score by backtesting against churned accounts from the previous 12 months — which signals went red first, and how far in advance of the churn decision?