CSM Comp Models Are Optimized for Churn Concentration, Not Customer Success

CSM compensation structures that weight expansion and net retention over retention incentivize portfolio concentration that masks underlying churn until it's too late. You end up with fewer, riskier customer relationships supporting the same revenue target.

The mechanism is simple. A CSM can hit a net retention target when a handful of high-spend customers expand aggressively while a quieter set of customers churns or contracts. The expansion moves the needle. The churn doesn't. So the comp model says the CSM performed well. The CSM performed well. The portfolio, though, is now structurally weaker: revenue is concentrated on fewer relationships, and the baseline cohort is eroding faster than expansion can paper over.

At portfolio scale, this problem compounds across the team. Everyone is optimizing their own net retention target independently. The aggregate effect is a customer base where revenue concentration increases every year, while churn on the non-expanded segment accelerates. It looks fine in the summary metrics. Cohort analysis reveals a ticking timer.

Why This Happens

The appeal of net-retention-based comp is obvious. Expansion revenue is harder to close than retention. It feels like it should be rewarded. So you make it a meaningful share of the CSM's variable comp and tie payouts to hitting a blended target. You're trying to activate your existing customer base as a growth lever. That's rational.

But you've just created an incentive for CSMs to concentrate effort on accounts that can expand, and to accept churn on accounts that won't. If a CSM has five accounts that could expand and ten that are at-risk for churn, the comp model says: fix the five. The ten are a cost center relative to your target.

Worse, you've created hidden flexibility. A CSM managing a double-digit book with a net retention target might actually be more effective with a smaller load: deeper engagement, fewer relationship breaks, higher retention. But if they can hit the blended target with the larger load, the system doesn't penalize the slack. In fact, it rewards it. More customers means more potential expansion levers. So CSMs tend to hold customer load longer than is operationally sound, because the marginal revenue on the expansion side outweighs the marginal risk on the retention side.

The Real Cost

The cost isn't visible in quarterly metrics. It's visible in cohort-level churn acceleration and in revenue concentration risk.

Run a cohort analysis and ask what share of current revenue is coming from long-tenured accounts. If that concentration keeps rising, it is not growth. It is churn being masked by expansion among the accounts that stayed. You're extracting more from fewer relationships. Eventually, those relationships terminate due to customer consolidation, technology shift, leadership change, or budget cycle. When they do, you don't have a broad base of healthy mid-market relationships to offset the loss. You have a portfolio optimized for extraction, not resilience.

CSM comp structures that separate expansion from retention also create timing problems. If a CSM is compensated primarily for net retention hits and secondarily for retention, they'll naturally front-load expansion conversations with at-risk accounts as a way to offset the churn risk to their comp. "Let's expand your contract in Q4 to secure your footprint for next year." Sometimes this is genuine product fit. Often, it's a hedge against a renewal that's actually slipping. The expansion masquerades as success. The underlying trend is degradation.

The Structural Pattern

The incentive misalignment concentrates across teams. In any CS org running net-retention-weighted comp at scale, cohort analysis will eventually show newer customer cohorts churning faster than older ones despite stable process and tooling. The deterioration is masked by expansion in surviving accounts. Net retention stays healthy because those accounts are expanding. The churned accounts disappear inside the blended number.

CSMs managing those cohorts are hitting their comp targets. They're earning bonuses. They're also, unknowingly, managing portfolios where underlying health is degrading and risk is concentrating. When the first major account announces churn, the scramble begins. But the preceding years weren't years of neglect. They were years of optimized incentive-following. The system worked as designed. The outcome was structural decline.

What's Actually Being Optimized For

CSM comp models aren't aligned to customer success. They're aligned to revenue extraction from customers who are already acquired and committed. That's a valid business need. But it's not CS. It's account management with a thinner margin and longer contract terms.

The problem isn't the CSMs. It's the design. You've created a system where retention without expansion is a penalty relative to the alternative (expansion with churn). The CSM's job is to optimize locally within those constraints. That's what incentive systems do. They optimize humans toward metrics. If the metric is blended, and the blended metric can be achieved multiple ways, the system doesn't distinguish.

What actually correlates with durable revenue at scale: retention of mid-market and lower-tier accounts (the majority of your base), expansion in accounts that organically need your product for new use cases (not forced), and expansion velocity that's sustainable across renewal cycles (not front-loaded before churn). None of these are singular metrics. All of them require CSMs to optimize for the hard part: keeping customers healthy and engaged across their business, not just extracting.

The Fix Is Structural, Not Tactical

You can't fix this with a tweak to the net-retention weighting. The problem is that you've separated expansion from retention and made them independently optimizable. A CSM can win on the blended number while losing on retention. That's not a weighting problem. That's a design problem.

The alternative is to tie comp to cohort-level outcomes. CSMs own a set of customers from day one. Their comp is tied to gross retention and net retention of that cohort measured annually and judged relative to the cohort's starting size. A CSM who keeps more of the cohort and expands appropriately should earn more than one who loses more customers but posts a prettier blended number. The difference forces the right trade-off.

This shifts the incentive from "expand or accept churn" to "keep customers healthy, and expand the ones who benefit." It also requires CSMs to be more selective about customer load, because underperforming a cohort metric is harder to hide with concentration.

It's also harder to execute. It requires patience with CSM comp resets. It requires belief that low-churn, moderate-expansion bases are more valuable than high-expansion, high-churn bases (they are, but it takes cohort data to prove it, and that takes three years). It requires CSMs to own accounts with full accountability, which means some of them will miss comp if they inherit struggling situations.

The current model is easier. It also builds portfolios that look better than they are, until suddenly they're not.