Most SaaS companies think their customer health problem is a data problem. They have too much data scattered across too many tools. The actual problem is synthesis. Nobody has assembled the signals they already have into one coherent health picture.
Product analytics shows usage. Support tickets show friction. The CSM knows which accounts feel risky from conversations. NPS shows what customers say when asked.
The contract value and renewal date sit in the CRM. Each of these signals is in a different system. No single team member can see all five at once.
The Five-Signal Customer Health Dashboard assembles these signals into one maintained view. This article is for the team where at-risk accounts are identified late because no single person had the full picture of each account’s health.
Why Customer Health Gets Identified Too Late
The obvious failure mode: a customer churns and the team says “we saw signals.” The signals were visible in product analytics. Nobody checked.
The CSM had noted low responsiveness in recent check-ins. The note was in the CRM, unread. The support ticket volume had doubled in the prior month. Nobody connected it to renewal risk.
The less visible failure is departmental ownership. Product analytics belongs to the product team. Support tickets belong to the support team.
CSM notes belong to customer success. NPS belongs to whoever runs the NPS survey. When health data is spread across owners with no synthesis layer, each owner sees their slice and nobody sees the whole.
The deepest failure is reactive intervention. The CSM who identifies a churning account two weeks before renewal is doing damage control. The CSM who identifies the same account two months before renewal has time to help. The synthesis layer that provides two months of warning is worth more than any intervention at the two-week mark.
The Five-Signal Customer Health Dashboard
Signal 1: Product usage
Product usage is the most objective health signal. Are the right users using the right features at the expected frequency?
The product usage signal captures:
- Login frequency over the last 30 days versus their baseline
- Key feature activation rates
- Team member adoption rate (are they using it individually or as a team?)
A sharp drop in product usage is one of the most reliable leading indicators of churn risk. It almost always precedes the customer’s conscious decision to churn.
Signal 2: Support interaction frequency and type
Support interaction signals the health of the customer’s experience. Rising ticket frequency indicates accumulating friction. Zero tickets over a long period could indicate successful self-service or could indicate disengagement.
The support signal captures:
- Ticket volume trend over 60 days
- Categories of tickets (onboarding friction vs. advanced feature requests vs. unresolved issues)
- Time to resolution
A customer with rising ticket frequency on unresolved issues is different from one with rising ticket frequency on new feature requests. The type matters as much as the volume.
Signal 3: CSM relationship notes
The CSM relationship note captures the qualitative health that the quantitative signals can’t surface. Responsiveness to emails. Engagement level in check-in calls.
Mentions of competitive alternatives. Organizational changes that affect the product champion.
The CSM relationship signal captures:
- Response time trend to check-in outreach
- Sentiment in recent conversations (notes flagged by the CSM)
- Organizational change flags (key champion leaving, department restructuring)
This signal requires CSM discipline to maintain. But when it’s maintained, it surfaces risk that product analytics alone cannot identify.
Signal 4: NPS and satisfaction
NPS data is a snapshot, not a trend. Its value in health monitoring comes from tracking the trend over time and combining it with the other signals.
A declining NPS trend across three consecutive surveys is a risk indicator. A low NPS score in combination with declining product usage and rising support tickets is a churn warning. A low NPS score with high product usage and low support tickets is a different kind of problem.
Signal 5: Commercial indicators
Commercial indicators are the business-side signals that suggest renewal health. Are they on the right plan for their usage? Are they in the right tier? Has usage expanded to the point where an expansion conversation is warranted?
The commercial signal captures:
- Days until renewal
- Usage relative to their current plan limits
- Expansion indicators (team growth, new use cases mentioned)
The commercial signal converts health data into actionable conversations: intervention for at-risk accounts, expansion for successful ones.
How the Living Library Updates the Health Dashboard
Monday morning, the Customer Health Signals Dashboard is current. Seventeen accounts are flagged: eight with decreased product usage over the past two weeks. Five have support ticket frequency above their baseline, and four have NPS responses below their established average.
Kiluma’s Living Library aggregates the five health signals from wherever they’re generated into one maintained picture. Product analytics, support ticket records, CSM notes, NPS responses, and contract data all flow into the Customer Health Collection. The dashboard updates on a schedule the team sets.
When a specific at-risk account needs investigation, the Conductor, Kiluma’s context-aware AI, surfaces the account’s full signal history. The CSM enters the intervention conversation with complete context: what the product usage looks like, what support issues are open. Also included: what was said in the last check-in, and what the renewal timeline is.
The onboarding process from Article 32 is the foundation the health monitoring builds on. The customer who completed onboarding milestones is monitored against a different baseline than the customer who didn’t.
Configure the Dashboard Before the Next Renewal Cycle
Before the next major renewal cycle, configure the five signals into one view. The product analytics platform, the support tool, the CSM notes, the NPS data, and the CRM all have API access or export capability.
The goal is not a perfect integration. The goal is one document or dashboard that a CSM can review in 10 minutes for any account. That view shows all five signals in one place.
The Health Signal That Gets Synthesized Is the One That Prevents Churn
The at-risk account that nobody caught is almost always the account where one team saw a signal and another team saw a signal. Nobody connected them. The Five-Signal Dashboard is the synthesis layer that connects them. Try Kiluma free for 14 days at kiluma.ai.