The manager who couldn't fire anyone

A manager I coached had given the same person the same piece of feedback in eleven monthly one-on-ones. Eleven. The performance hadn't moved. The conversation had become a ritual: a polite restatement of expectations, a polite acknowledgment, a polite return to the same behavior the following week. The manager described the cadence as working on it. The HR partner described it the same way. The senior leadership team had not asked the question in two quarters.

He told me he was being patient. The team, when I talked to them, used a different word.

The team had learned something the manager did not realize he was teaching. They had learned the stated standard was decorative. The real standard — the one they could see in the manager's actual behavior month over month — was whatever this person was doing. Three of the strongest engineers had quietly stopped pushing on the things the manager kept asking for in standups, because there was no observable difference between doing them and not. Excellence had been informally repriced as optional. The repricing was invisible in the dashboards and visible in everything else — the small signals of who showed up to discretionary work, who volunteered for the hard projects, who took the late-night escalation without being asked.

This is the hidden math of the manager who can't end a tenure. The cost is not paid by the manager. The cost is paid by the four or five people on the team who recalibrate downward over the following two quarters, because nothing in the environment rewards them for staying high. By the time the manager finally acts, the team he had a year ago is no longer the team he has. The strong ones got quieter. Some left.

Patience is the right word for the first three conversations. The first conversation surfaces the issue. The second confirms whether the surfacing produced any change. The third tests whether sustained attention produces the change. By the fourth, the same conversation is no longer patience. It is avoidance dressed in HR vocabulary. The question is not are we being fair to this person. It is what are we teaching the other nine people in this room by waiting.

The avoidance has predictable causes. The manager has not yet had a difficult termination conversation, and the prospect of having one is more frightening than the cost of not having one. The HR partner, optimizing against legal risk, is recommending more documentation rather than action. The leadership team above the manager is graded on retention numbers, which biases against any termination decision. Each of these incentives is individually defensible. The aggregate is the eleven-month performance loop with no resolution.

The strong performers read the loop accurately. They notice that the company has stated standards and tolerates a person who fails them. They calibrate their own behavior against what the company actually rewards, which is observable in the loop, not in the standards document. The calibration produces a team that is, after two quarters, operating at the lowest tolerated standard rather than at the standard the manager has been articulating. The manager has, by inaction, redefined the team's operating norm.

The damage is asymmetric. The manager believes he is preserving the team by being patient with the underperformer. The team experiences the same patience as evidence that effort is not noticed. The underperformer is not the cost the manager thinks he is preserving against; the cost is the strong performers who quietly recalibrate, and that cost is invisible until two of them resign in the same fortnight.

The manager finally let the person go in the fifteenth month. The termination conversation was brief. The team's response was uncomfortably positive — too positive, in a way that signaled how much the team had been carrying. Two of the strong engineers told me, separately and without prompting, that they had been three weeks from leaving. They stayed because the firing told them the standard was real after all.

The cost of the fifteen-month delay is impossible to recover. The strong engineers who had recalibrated downward during the loop did not snap back to their previous effort level the moment the underperformer left. The norms had shifted. The team's operating standard had repriced. Restoring the standard required the manager to actively re-set expectations in the months after the termination, against a team that had been trained to expect the old, lower bar. The work was meaningful, took two quarters, and was successful only because the manager had finally acted before the team had completely lost faith.

Most managers in this position do not act before the team loses faith. The eleven months become eighteen, then twenty-four. By month twenty-four, the strong performers have either left or have permanently calibrated to the lower bar. The team is now a different team than the one the manager was preserving. The eventual termination, when it comes, no longer restores anything. The damage is structural.

The hardest hire you will ever make is the one you should have made a year ago by ending someone else's tenure. Until you make it, the team is running on borrowed conviction. The loan comes due quietly, and the interest is paid in your best people leaving without telling you why.

Before your next monthly review with an underperformer, ask:

• How many times have I delivered this same feedback, with no measurable change in behavior?
• What specifically would need to change for the next conversation to be different from the previous three?
• What are the strong performers on this team learning from watching me have this conversation repeatedly?
• If this person were not on the team and we were considering hiring them today against the role's current requirements, would we hire them?

The honest answers point at the conversation the manager has been postponing. The conversation costs one uncomfortable hour. The deferral costs the strong performers who quietly stop trying. The math is consistently unflattering to the deferral, and most managers run it too late.