Heroes don't scale. Processes do.

Every CS team has one. The CSM who remembers every account's history, catches the risk everyone else misses, and stays late to fix the renewal nobody saw coming. Leadership calls this person a top performer. It's more accurate to call them a single point of failure.

Heroics work until they don't — until that person takes leave, moves teams, or simply runs out of hours. What they were doing wasn't a skill unique to them. It was a process nobody had bothered to write down.

The fix isn't asking everyone else to work as hard. It's extracting what the hero actually does — the sequence of checks, the questions they ask, the moment they escalate — into something a newer CSM could follow on their second week. A playbook, not a personality.

This is uncomfortable for the hero. Writing the process down feels like giving away the thing that makes them valuable. In practice it's the opposite: it's what makes their judgment reusable instead of trapped in one person's calendar.

Two questions test whether a CS function has actually done this: if your best CSM left tomorrow, would the accounts they hold be fine in a month? And can a new hire find the answer to a routine question without asking a person?

If the answer to either is no, the function is still running on heroics — however good the results look this quarter.