Home / Blog / The Churn Your Health Score Can’t See

header for The Churn Your Health Score Can’t See

The Churn Your Health Score Can’t See

Renewal conversations rarely start at renewal time. That’s the uncomfortable truth Customer Success leaders across Europe have been wrestling with, and it’s reshaping how they approach churn risk. The churn risk indicators were there weeks, sometimes months, before the conversation turned difficult. The problem was knowing what to look for, and having the organizational infrastructure to act on it.

This is where Post-Sales is being pushed to mature: not just tracking health scores, but building the discipline, the cross-functional habits, and the signal literacy to intervene before a customer has already made a decision.


Most Teams Track Churn Indicators. Most Miss the Ones That Matter.

Most Post-Sales teams today operate with some version of a health score as their primary churn risk indicator. Usage data, NPS results, contract dates, and payment history are the common ingredients. These scores give teams a starting point. But across discussions in Munich, Ghent, Hamburg, Berlin, and Paris, one observation kept surfacing: the score is green, and the customer is still at risk.

Health scores measure what happened. They capture product usage patterns from last week, NPS scores submitted last quarter, and renewal dates still months away. What they don’t capture is what a CSM notices in a conversation: the hesitation before answering a question about ROI, the shortened meeting that used to run an hour, the executive sponsor who has stopped joining calls.

In a Ghent discussion moderated by Linda Tilkin, participants put it directly. A customer can have green metrics and still face high churn risk if they don’t feel progress is happening. That framing, churn as a perception and context problem rather than a data problem, cuts to the heart of what CS teams get wrong when they treat health scores as the final word on account health.

Several leaders across discussions noted that their teams struggle to agree on what “healthy” even means. You can build a health score before you’ve defined what a healthy customer looks like. But you’re then measuring the absence of problems rather than the presence of success. That distinction matters when prioritising risk, and most teams haven’t fully made it yet.


Where Better Signal Detection Is Making a Difference

Enriching What Goes Into the Score

The teams making real progress aren’t abandoning health scores. They’re expanding what goes into them. Beyond usage and NPS, several groups described adding event attendance, email engagement rates, CSM sentiment ratings, and early indicators of financial pressure flagged during check-in conversations.

In a discussion moderated by Amy Downs in Hamburg, participants described a four-field model built in Salesforce: usage, NPS, contract renewal proximity, and payment behavior. A fifth field captures CSM sentiment, entered manually and scored weekly. The manual entry creates a disciplined team conversation about where accounts actually stand, separate from what the system shows. Teams using this hybrid model reported better visibility into accounts that were quietly drifting, even while their usage remained relatively stable.

The data on onboarding is stark. Teams that tracked onboarding completion as a formal churn risk indicator found a clear pattern. Customers who didn’t complete structured onboarding churned at a 60% higher rate than those who did. That turns onboarding from an engagement activity into a hard risk flag, one that needs monitoring from the first weeks of the customer relationship, not the last quarter before renewal.

Reading the Signals Dashboards Can’t See

Participants across multiple cities pointed to a consistent set of early warning signals that don’t appear in any system. Champion ghosting (when a key internal contact stops responding, reduces attendance, or is simply no longer the right person in the room) emerged as one of the most reliable churn risk indicators. Competitive mentions during routine calls, changes in who joins customer meetings, data export requests, and unusually long approval cycles for minor decisions all surfaced as signals experienced CSMs have learned to watch.

In another discussion moderated by Giorgia Pedenzini in Munich, participants described a category of events that essentially override all other health indicators. Mergers and acquisitions sit at the top of that list: when a customer goes through an M&A event, historical data from these teams suggests a churn probability of 95 to 98%. That’s a signal requiring immediate escalation to leadership, not a flag to sit in a CSM’s weekly review. Three-scenario forecasting, which tracks best-case, worst-case, and most likely outcomes simultaneously for at-risk accounts, helped some of the same teams improve renewal forecast accuracy to around 85%. The method works by forcing honest internal conversations about where accounts actually stand, rather than defaulting to optimistic assumptions.


What’s Slowing Progress

The Data Problem

The most meaningful risk signals are also the hardest to operationalize. Relationship health, stakeholder sentiment, and organizational changes inside a customer’s company all require human observation to detect early. AI-assisted sentiment analysis is being explored in some teams as a way to surface patterns from call notes and email threads, but most teams are still experimenting. The quality of what gets surfaced depends heavily on whether CSMs are consistently capturing qualitative notes in a structured way, which is itself an adoption challenge most teams haven’t fully solved.

Data silos compound the problem. CS teams frequently lack visibility into signals that live in other systems: support ticket sentiment in one platform, sales activity in another, product telemetry somewhere else. Even teams with sophisticated CS platforms described spending significant time reconciling data manually before they could see a complete account picture.

The Coordination Problem

Nearly every discussion surfaced some version of the same structural challenge: CS identifies the risk, but CS can’t resolve it alone. Rescue plans require Sales to re-engage on positioning and value, Product to communicate roadmap items, and leadership to show up in executive conversations. Coordinating that response before a customer has formally raised concerns is when intervention is most likely to work. But most organisations lack a clear trigger for involving other teams. The flag tends to go up only after the situation is already serious.

The Sales-CS handoff remains a persistent source of early risk. What gets promised during the sales process shapes the churn risk indicators that emerge later. When expectations about product capability, implementation scope, or stakeholder alignment are off from the start, CS inherits the gap.


Where CS Leaders Are Really Uncertain

Automation vs. Judgment

The appeal of automating risk detection is real. It removes inconsistency in how risk is spotted across a team and creates a traceable record of when signals were flagged and what happened next. But the signals with the highest predictive value (tone, silence, the sense that a relationship has shifted) are precisely the ones hardest to automate. Several leaders expressed concern that over-indexing on what the system can measure leads to under-valuing what an experienced CSM can observe.

Interestingly, not everyone agreed that human judgment should function as a supplement to data. Some participants argued for the reverse: that a health score should primarily exist to surface accounts that deserve closer human attention, rather than serve as a verdict on account health in its own right.

Lean Models vs. Comprehensive Ones

There’s genuine tension between building health scores that capture every relevant dimension and keeping them simple enough that teams actually trust and use them. The argument for lean models is strong. A score tracking fifteen variables across multiple systems becomes impossible to explain to a customer-facing team. The weights quickly become arbitrary. The argument for richer models is equally strong. A score built on two or three inputs will routinely miss accounts that look fine but aren’t.

None of the discussions resolved this cleanly. The direction with the most consensus: start lean. Validate against actual churn outcomes. Add factors where the data shows they improve prediction. Remove those that turn out to be noise. It’s a discipline of iteration, not a design problem with a single right answer.


What Changes When CS Becomes a Risk Governance Function

These conversations collectively point toward a shift in what CS is being asked to do. The role is moving from reactive retention (respond when risk is flagged, save what you can) toward structured risk governance. That requires different skills, different rhythms, and a different level of cross-functional buy-in.

CSMs are increasingly expected to function as analysts as well as relationship managers. That means reading patterns across accounts, recognising when a signal warrants escalation, and framing risk in terms that Sales and leadership will act on. Discussions in Paris emphasised a key shift: Post-Sales needs to stop treating churn prevention as its sole responsibility. Holding other functions accountable for the responses they need to provide is increasingly part of the job.

As AI tools mature, they take on more of the pattern recognition and administrative work that currently occupies CSM bandwidth. That frees up space for the distinctively human parts of the role: discovery conversations, executive relationships, and navigating organisational complexity inside the customer. Those capabilities become more valuable, not less. Participants in Dublin noted that the differentiator in Post-Sales will increasingly be the quality of judgment and the depth of the relationship, not the efficiency of the process.


What CS Leaders Are Experimenting With

Several concrete practices emerged across discussions in multiple cities.

Rescue Workflows and Learning Rituals

Teams in Hamburg have built mandatory win-back checklists that CSMs must complete before escalating a churning account to senior leadership. The checklist ensures CSMs document and attempt specific interventions first: additional training, an executive alignment call, and an ROI workshop. When a CSM completes the checklist without resolution, the company offers a direct CEO Q&A as a final retention lever.

Several teams are running regular post-mortems on churned accounts as a forward-looking exercise rather than a retrospective one. The review maps what signals were visible and when, what actions were taken, and what might have changed the outcome. When done consistently, this builds collective signal literacy across the Post-Sales team in ways that no training program can replicate.

Expanding the Signal Detection Net

In a discussion moderated by Caroline Bresson in Paris, participants noted that community events and networking gatherings have become useful informal detection channels. Customers share concerns more openly when conversations aren’t structured around product topics. CSMs who attend these events regularly report catching early churn risk indicators that would never surface in a scheduled check-in.

Paris-based teams are also embedding Product Owners in strategic customer meetings to close the loop between customer experience and product priorities. The practice signals to customers that they’re being heard beyond the Customer Success relationship, which reduces one category of perceived risk.

Some Munich teams track three scenarios weekly for at-risk renewals: best case, worst case, and most likely. The practice has improved forecast accuracy and shifted internal conversations toward which accounts actually need intervention, versus which ones appear concerning in a single snapshot.

Meanwhile, teams in Amsterdam described using a structured Identify-Diagnose-Act-Monitor cycle for at-risk accounts. The cycle separates diagnosing the root cause (through open questions) from acting on it (through a joint success plan). Teams that skip the diagnosis step tend to apply the wrong intervention.


Three Priorities for CS Leaders Building Risk Capability

The clearest implication from these discussions: building risk management capability is a long-term organisational project, not a technology purchase. A new CS platform won’t solve the problem on its own. Teams still need to define what a healthy customer looks like, establish rhythms to consistently review risk, and do the internal work to make churn prevention a shared responsibility.

Define Success Before You Measure Risk

Defining success before measuring risk is where most teams should start. What does a healthy customer look like twelve months in? What patterns of usage, engagement, and relationship depth correlate with expansion and advocacy? Building the health score from that definition outwards produces something useful. Assembling a list of negative indicators and labelling it a model does not.

Build Cross-Functional Accountability

Making the case for cross-functional ownership is the second priority, and it requires Post-Sales leaders to reframe the conversation. Risk management isn’t a Customer Success problem that Customer Success is asking for help with. It’s a revenue problem that Sales, Product, and Operations share accountability for. Making that case persistently, with data, is increasingly part of what senior Post-Sales leadership needs to do.

Establish the Post-Mortem Ritual

Systematically reviewing churned accounts builds the signal literacy that no platform provides. The goal isn’t to assign blame. It’s to build pattern recognition. The teams with the strongest churn risk indicator capability aren’t distinguished by their tools. They’re distinguished by how consistently they learn from what they got wrong.


The shift from reactive retention to proactive risk governance doesn’t happen because a team buys a better tool or builds a more sophisticated health score. It happens because Post-Sales leaders decide to treat risk management as a discipline. That means building its own rhythms, developing its own language, and establishing its own cross-functional demands.

The teams making the most progress on churn prevention aren’t the ones with the best data. They’re the ones that have built the habits, the internal credibility, and the organisational alignment to act on what they already know. A health score can tell you where to look. What you do with that information depends on everything else.

Post-Sales is moving toward a model where churn is neither a surprise nor a Customer Success department problem. Getting there requires building the infrastructure and the culture for proactive risk management. That means designing for the accounts that look fine until they suddenly aren’t, not just the ones that are obviously in trouble.


Join the Conversation

The discussions in this article came from Post-Sales professionals meeting in person across Europe. If you want to be part of the next one, find an upcoming event near you and confirm your spot. Can’t make it in person? Join the Digital Snacks membership to access our virtual sessions and connect with the European Post-Sales community wherever you are.

Find your hyper-local community

Find your local community in Dublin, Amsterdam, Stockholm, Malmo, Copenhagen, Berlin, Munich, Hamburg, Ghent, Paris, Madrid, Zurich, Bucharest and more.

Get invited to Events, Job alerts & More. Straight to your inbox.

close-link