Customer Health Score: What It Measures and Why It Matters
A customer health score is a composite metric that tells you how likely a customer is to stay, expand, or churn, based on a combination of product usage, support interactions, payment behaviour, and engagement signals. It takes data that lives in separate systems and turns it into a single, actionable number that revenue teams can act on before a problem becomes a cancellation.
Done well, it shifts your retention work from reactive to predictive. Done badly, it becomes a dashboard metric that nobody trusts and nobody uses.
Key Takeaways
- A customer health score is only as useful as the signals you choose to include. Vanity inputs produce vanity outputs.
- Most churn is visible in the data weeks before the customer says anything. Health scoring makes that window actionable.
- The biggest implementation failure is building a score nobody owns. Without a clear owner and a clear response protocol, the number is decorative.
- Health scores work best when they are tied directly to a playbook: green means this, amber means that, red means call them today.
- Marketing’s role in health scoring is often underestimated. Engagement with your content, campaigns, and community is a legitimate leading indicator of retention risk.
In This Article
- Why Most Companies Build Health Scores That Don’t Work
- What Should Actually Go Into a Customer Health Score
- Product and Service Engagement
- Support and Satisfaction Signals
- Commercial and Relationship Signals
- Marketing Engagement as a Health Signal
- How to Weight the Inputs
- Turning a Score Into an Action
- Who Owns the Customer Health Score
- The Measurement Problem
- Building the Score Incrementally
- Health Scoring in the Context of Go-To-Market Strategy
- The Honest Limitation
Why Most Companies Build Health Scores That Don’t Work
I have seen this pattern more times than I can count. A company decides it needs a customer health score, assigns someone in operations or customer success to build it, and three months later there is a colour-coded dashboard that the sales team ignores and the CS team checks once a week. The score exists. It does nothing.
The failure is almost never technical. It is structural. The score was built without agreement on what “healthy” actually means for that specific business. Someone picked inputs that were easy to pull from the CRM rather than inputs that genuinely predict retention. The thresholds were set by instinct rather than by looking at what the data actually showed for customers who churned versus customers who renewed.
When I was running an agency and we started losing clients we thought were happy, the honest post-mortem was always the same: the signals were there. Declining email open rates on our reports. Fewer stakeholders attending QBRs. Slower response times on approvals. We were not measuring any of it. We were relying on the account manager’s gut feeling, which is a terrible instrument for early warning.
A well-constructed health score is essentially a formal version of what a great account manager does intuitively. It reads the room at scale.
What Should Actually Go Into a Customer Health Score
There is no universal formula. The inputs that predict churn in a SaaS product with daily active users are different from the inputs that predict churn in a professional services relationship or a B2B subscription with quarterly touchpoints. The starting point is not “what data do we have” but “what behaviour did our churned customers exhibit that our retained customers did not.”
That said, there are categories of input that tend to matter across most business models.
Product and Service Engagement
For software businesses, this is login frequency, feature adoption, time in product, and whether the customer is using the parts of the product that correlate with long-term value. A customer who signed up for your platform but only ever uses one of five core features is a different risk profile from one who has integrated it into their daily workflow.
For services businesses, the equivalent signals are meeting attendance, stakeholder engagement, responsiveness, and whether the relationship has grown beyond the original point of contact. Single-threaded relationships are fragile. If your only champion leaves, you are starting from scratch.
Support and Satisfaction Signals
Ticket volume is a double-edged signal. A customer who raises no support tickets could be perfectly happy or completely disengaged. A customer who raises a lot of tickets could be frustrated or could be deeply invested in getting value from your product. The nature of the tickets matters more than the volume.
Unresolved issues, escalations, and negative satisfaction scores are cleaner inputs. NPS and CSAT have their limitations as standalone metrics, but as one component in a composite score they add something. A customer who scores you a 3 out of 10 three months before renewal is telling you something important.
Commercial and Relationship Signals
Payment behaviour is one of the most underused inputs. Late payments, disputes, and downgrade requests are almost always leading indicators of churn, not lagging ones. A customer who starts paying late is often a customer who is internally questioning the relationship before they have said anything to you.
Contract and renewal engagement matters too. How far in advance did they engage on renewal last time? Did they ask for a price reduction? Did they request a shorter contract term? These are signals worth capturing.
Marketing Engagement as a Health Signal
This is the one that marketing teams consistently underestimate. A customer who opens your product updates, attends your webinars, and engages with your community is a customer who is invested in the relationship. A customer who has not opened a single communication in four months is not necessarily churning, but they are drifting.
Email engagement data, event attendance, content consumption, and community participation are all legitimate health inputs. The go-to-market thinking that treats marketing as purely an acquisition function misses this entirely. Post-sale marketing engagement is a retention signal. If you are building a growth strategy that ignores what happens after the contract is signed, you are working with an incomplete picture. The broader principles behind that kind of commercial thinking are covered in the Go-To-Market and Growth Strategy hub.
Vidyard’s work on why go-to-market feels harder touches on exactly this point. Buyers are more informed, more sceptical, and more likely to disengage quietly than they used to be. That applies to existing customers as much as prospects.
How to Weight the Inputs
Once you have agreed on your inputs, you need to weight them. Not all signals carry equal predictive value. In most SaaS businesses, product usage tends to be the strongest predictor of retention. In professional services, relationship depth and stakeholder engagement tend to dominate. In high-touch B2B, commercial signals often matter most.
The honest answer is that weighting should be informed by your own historical data, not by what someone else’s framework suggests. If you have two years of customer data and you can look back at the cohort that churned, you can run even a simple analysis to see which signals were most divergent between churned and retained customers. That is your weighting guide.
If you do not have that data yet, start with a hypothesis-based model and commit to refining it over two or three quarters as you accumulate evidence. A rough model that gets iterated is more valuable than a perfect model that never gets built.
BCG’s work on commercial transformation is worth reading in this context. The core argument, that commercial effectiveness comes from better targeting and better prioritisation rather than just more activity, applies directly to how you allocate retention effort based on health score data.
Turning a Score Into an Action
This is where most implementations fall apart. The score exists. The thresholds are set. Green, amber, red. And then nothing happens, because nobody agreed in advance what “red” actually means in terms of who does what by when.
A health score without a playbook is a weather forecast you cannot act on. You need to define, in advance, what each threshold triggers. A customer moving from green to amber might trigger an automated check-in sequence. A customer hitting red might trigger an immediate call from the account manager or CS lead. A customer who has been red for thirty days without improvement might trigger an executive escalation.
The playbook also needs to address the other end of the spectrum. Customers who are scoring consistently green are your expansion candidates. They are the ones who are getting genuine value and are most likely to respond positively to an upsell or cross-sell conversation. Health scoring is not just a churn prevention tool. It is a revenue intelligence tool.
I spent time working across businesses where the sales and marketing teams were almost entirely focused on new logo acquisition. The existing customer base was treated as a maintenance problem rather than a growth asset. When we started mapping customer health properly, we found a meaningful cohort of high-engagement, high-satisfaction customers who had never been approached about additional products or services. The revenue was sitting there. We just had not looked.
Who Owns the Customer Health Score
Ownership is a political question as much as an operational one, and it is one that teams tend to avoid until it causes a problem. In most businesses, the health score sits with customer success. That is the right default. CS teams have the most direct relationship with customers and the most context for interpreting the signals.
But marketing needs a seat at the table. Marketing owns several of the input signals, particularly around engagement and communication. Marketing also owns the content and campaigns that can move a customer from amber back to green. If marketing is not connected to the health scoring system, you are leaving a significant lever unpulled.
Sales needs visibility too, particularly on expansion opportunities. A health score that is only visible to CS and invisible to the account executives responsible for renewals and upsells is a broken system.
The practical answer is that the score should live in a shared system, with clear ownership of each input category and clear accountability for each response playbook. It is a cross-functional instrument, even if CS is the primary operator.
The Measurement Problem
Health scores are imperfect by design. They are a model of customer reality, not reality itself. A customer can score green and still churn because their internal champion left, their budget was cut, or a competitor made them an offer that had nothing to do with their satisfaction with your product. A customer can score amber and renew because the relationship at the executive level is strong in ways your data does not capture.
This is not an argument against health scoring. It is an argument for treating the score as one input into a human decision, not as an autonomous verdict. I have always been sceptical of the idea that dashboards replace judgment. They inform judgment. The best account managers I have worked with used data to sharpen their instincts, not to replace them.
The score is most valuable as a prioritisation tool. It tells you where to focus limited CS and account management capacity. It does not tell you what to say when you get there. That still requires context, empathy, and commercial intelligence that no algorithm captures cleanly.
Building the Score Incrementally
One of the more useful pieces of advice I have given to teams starting this work is to resist the temptation to build the perfect model before launching anything. A five-signal health score that you can build in a month and start acting on is worth more than a fifteen-signal model that takes six months to instrument and never quite gets finished.
Start with the three or four signals you are most confident about, the ones where you have clean data and a reasonable hypothesis about their predictive value. Build the playbook around those. Run it for a quarter. Look at what the score predicted versus what actually happened. Refine the weights. Add signals where you see gaps. Iterate.
Tools like Hotjar can contribute to this picture on the product engagement side, capturing how customers interact with web-based products and surfacing friction points that might not show up in support tickets. The Hotjar platform is one of several options for capturing behavioural signals that feed into a broader health picture.
The growth hacking literature has some useful frameworks for thinking about which metrics actually predict retention versus which ones just feel important. The Semrush breakdown of growth hacking examples covers several cases where companies identified the specific engagement signal that correlated with long-term retention, and built their entire activation strategy around driving customers to that moment as fast as possible. The same logic applies to health scoring: find the signals that actually predict the outcome you care about, and weight them accordingly.
Health Scoring in the Context of Go-To-Market Strategy
Customer health scoring is not a customer success tool that happens to sit near marketing. It is a go-to-market instrument. It tells you which customers are ready to grow, which are at risk, and where your post-sale investment should go. That is a strategic question, not an operational one.
The businesses I have seen grow most efficiently are the ones that treat their existing customer base as a primary growth asset rather than a baseline to defend. That shift in orientation changes everything: how you resource CS, how you structure marketing, how you think about the relationship between acquisition cost and lifetime value. Health scoring is the mechanism that makes that orientation operational.
Vidyard’s research into untapped pipeline and revenue potential for go-to-market teams makes a related point: a significant share of revenue potential sits in existing accounts that are not being worked effectively. Health scoring is one of the primary tools for identifying and capturing that potential.
If you are thinking about where health scoring fits within a broader commercial growth framework, the Go-To-Market and Growth Strategy hub covers the surrounding territory in more depth, from how to structure your market approach to how retention and expansion fit into a sustainable growth model.
BCG’s thinking on go-to-market strategy emphasises that commercial success depends on clarity about where you are playing and how you are winning. That principle applies as much to retention strategy as it does to launch strategy. Knowing which customers to invest in, and which to let go gracefully, is a strategic decision that health scoring data makes possible.
The Honest Limitation
I want to be direct about something that often gets glossed over in the health scoring conversation. No scoring model compensates for a product or service that is not delivering genuine value. If customers are churning because what you sell does not do what they need it to do, a better health score will not fix that. It will just give you earlier visibility into a structural problem.
Marketing and customer success can do a lot. They cannot make a bad product good. The businesses I have seen invest heavily in retention infrastructure without addressing underlying product or service quality issues tend to end up with very sophisticated early warning systems and the same churn rate. The score is a diagnostic. The treatment has to address the actual cause.
Used honestly, health scoring surfaces those structural issues faster. If you are seeing consistent red signals around a particular feature, a particular customer segment, or a particular use case, that is product feedback as much as it is a retention alert. The best teams use health data to inform product and service decisions, not just CS workflows.
About the Author
Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.
