Paying CSMs on Net Retention Without Punishing a Hard Book

A CSM role with heavy variable pay tied to retention starts to behave like a sales role, which changes who applies for it and how they act under pressure on a struggling account. A base-heavy structure with a smaller retention-linked pool keeps the role's actual job, proactive account health work, in focus, while still rewarding the outcome that matters. Weighting too far toward variable pay tends to produce CSMs who avoid flagging risk early because a red account threatens their own paycheck.

A CSM holding several accounts that were already troubled at handoff is not doing a worse job than one holding a book of healthy, well-onboarded accounts, even if the second CSM's net retention number looks better on a leaderboard. Rate each account's starting health at handoff, and measure the CSM against the change from that starting point, not against an absolute retention target that ignores what they inherited.

A CSM can influence onboarding quality, proactive check-ins, and how quickly they escalate a risk signal. They usually cannot control a pricing decision made by finance, a champion who leaves for reasons unrelated to the product, or a company-wide budget freeze. Build the comp plan around the first category and treat the second as a rated exception in reviewing performance, rather than holding the CSM fully accountable for outcomes outside their control.

A few structures show up repeatedly, each with a specific failure mode worth planning around.

• Straight percentage of book net retention: rewards whoever has the easiest book, as described above.
• Bonus for accounts hitting a health-score threshold: works well until the health score itself gets gamed to look better than reality.
• Team-wide pool split evenly: removes individual incentive problems but can demotivate a CSM carrying a disproportionately hard set of accounts.
• Milestone-based bonuses tied to specific account actions: aligns incentives closely but takes real setup work to define fairly across different account types.

A comp plan can hit its target retention number while quietly encouraging a behavior you did not intend, like a CSM delaying a churn-risk escalation because reporting it early affects a quarterly number they get measured on. Periodically ask CSMs directly whether the plan is changing how they handle a difficult account, and treat a pattern of delayed escalation as a plan design problem, not an individual performance issue.

Changing a comp plan CSMs have already built their expectations around, especially one that normalizes for book difficulty for the first time, needs a longer runway than most other policy changes. Share the new structure and the reasoning behind the normalization a full cycle before it takes effect, and walk each CSM through what it would have meant for their own last period's number, so nobody feels the change was designed around a specific person's book after the fact.

Without book-difficulty normalization, leadership loses the ability to tell whether a low retention number reflects a CSM problem or an account problem, which means every difficult conversation about performance starts from a weaker footing. A normalized number lets you say, specifically, that a CSM underperformed relative to what a comparable book should have delivered, which is a far more defensible basis for a performance conversation than a raw retention figure that never accounted for what they were handed.