Addressing poor performance without demotivating

Someone on your team is consistently late to the weekly meeting. Everyone has noticed. Nobody says anything. Neither have you, for four weeks now — and every week that passes makes the conversation harder to open.

It is the most ordinary situation in management, and one of the most poorly handled. Not for lack of courage: out of a reasonable fear of breaking something.

What waiting costs

Waiting feels prudent. It is in fact the most expensive decision available, for three reasons that compound.

The behaviour becomes the norm. A deviation tolerated for a month is no longer a deviation: it is an unwritten rule. Correcting it later looks like changing the rules midway, which is a far harder thing to have accepted.

The team is watching you. Your other reports see the lateness, and they see your silence. They draw conclusions about what is allowed. The cost never lands on one person only.

Your irritation accumulates. And on the day you finally raise it, the intensity of your reaction is out of proportion with that day's event. The other person receives four weeks of frustration for ten minutes of lateness, and finds it unfair. They are right.

Correction is not punishment

A corrective conversation is not a sanction. It is a restatement of what is expected, addressed to someone you assume will meet it. That assumption changes everything about how you run the exchange.

In practice it rules out three things: the accusatory past tense (“you've done it again…”), summoning witnesses (“everyone has noticed…”), and announcing a consequence before you have heard the answer.

Running the conversation

Open on the fact, not the intention

“You arrived after the start of the last four team meetings” is an observation. “You don't take these meetings seriously” assumes a motive — and you do not know whether it is true.

This distinction is not diplomacy. It is accuracy: you know the arrival times, you do not know the reasons.

Ask before you explain

One open question — “what's going on?” — before any explanation of yours. In a striking share of cases the answer teaches you something: a childcare constraint, a client call that always overruns, a misunderstanding about the real start time.

Had you opened with your expectations, you would have corrected the wrong problem.

Name the impact on the group

This is what makes the expectation legitimate. Lateness is not serious in itself; it is serious because it forces the team to revisit what has already been covered, because it shifts the meeting for eight people, because it establishes the idea that start times are advisory.

Close on something concrete

An explicit expectation, a date, and a check-in. Without a check-in, you will be reopening the subject from scratch in three weeks.

When it does not go smoothly

The sequence above assumes someone receptive. In practice you will meet three harder reactions.

The endless justification. Every instance is explained by an external circumstance. Do not argue each circumstance: return to the pattern. “I hear that there was a reason each time. I'm still looking at four out of four.”

The deflection. “What about so-and-so, who always leaves early?” One sentence handles it: “We can talk about that separately if you want. Right now I'm talking about you.” Then return to the subject. Do not get drawn in.

The silence. The most unsettling of the three. Let it run longer than you want to, then ask a closed question that requires an answer: “does what I'm asking seem workable to you?”

Why these conversations are hard to prepare alone

Re-reading a method the night before does not prepare you for those three reactions. What is really at stake in a corrective conversation is reflexes: holding your line while someone justifies themselves, not apologising for having standards, knowing when to stop once the essential has been said.

Those reflexes are built by repetition — which is exactly what you cannot do on your real team, where every failed attempt leaves a mark.