Head of Business
The Operator's Playbook

We manage on attrition, and high attrition is treated — automatically — as a problem. But a single churn number is a lazy measure, because it hides the most important distinction a leader can draw. There are two kinds of attrition, and they mean opposite things.

Here's why this matters, and why it's the kind of thing figures destroy: a raw churn number can't tell these two apart. So the board sees "big churn — that's a problem," and the manager, under pressure to make the number look better, does the easy thing: they retain the detractor to keep headcount stable. A plaster. Because managing the number is easier than fixing the culture — and fixing the culture means confronting behaviour and driving change, which is hard and uncomfortable and slow. So people manage the number instead, and in doing so they protect the very person damaging the department, because losing them would "look bad." The easy path quietly optimises for the wrong thing.

The right path is harder. You don't move someone to the right behaviour by tolerating the wrong one — that's the plaster. You address the misalignment directly: name the behaviour, connect it to the values and its impact on others, and give them a real, supported chance to change. And here's the safeguard that keeps this honest — the detractor is defined by conduct, not by disagreement. The person who challenges you well, who pushes back for good reasons, who's struggling because their onboarding failed them — none of those is a detractor. A detractor is someone whose behaviour corrodes the people around them, and who, given genuine support and a clear mirror, still won't align. You earn the right to call their departure non-regretted only after you've honestly tried to bring them with you. Do that, and letting them go isn't a failure of retention — it's an act of care for everyone who stays.

From experience — because I've paid for this one

I'm not writing this from theory. I've had the attrition I regretted — good people I fought to keep and lost anyway, to circumstances I couldn't beat. And I've managed the behaviours: I've sat in the hard conversations, held the line on values, and yes, I've had people leave because of it. It is genuinely difficult to replace experience — I won't pretend that cost isn't real, or that the pressure doesn't land when the churn number climbs and someone above you wants it explained. But you do the right thing anyway. Not because it's rewarded — often it isn't, and the figure looks worse before it looks better — but because protecting the culture for the people who stay is the job. The outcome and the pressure don't change what's right. They just test whether you'll do it.