Why rehearsing a conversation works

Managers prepare for difficult conversations by thinking about them. Usually in the shower, usually repeatedly, usually for several days. Then the conversation happens and the prepared version does not come out. This is not a failure of preparation; it is preparation of the wrong kind.

What is the difference between thinking it through and rehearsing?

Thinking rehearses the argument. Speaking rehearses the delivery, and the delivery is what fails.

When you plan in your head, you work with the gist. You know you will explain that the reports have been thin, that this has happened twice, and that you need it fixed. All of that is clear and none of it is a sentence. In the room you have to produce actual words in real time while managing your own discomfort and watching someone's face, and the gist does not survive that.

Saying it out loud once forces the gist into a sentence. That is a different cognitive task, and it is the one you will have to perform.

What actually changes when you say it aloud?

Three things, and the first is the most useful.

You find the sentence you cannot say. Almost everyone has one — usually the sentence that states the problem plainly. It comes out mangled, or hedged, or three times longer than intended. Finding it in your kitchen costs nothing. Finding it in the meeting costs the conversation.

Your opening gets shorter. Spoken aloud, a two-minute preamble is obviously a two-minute preamble. On the page or in your head it feels like appropriate context. Saying it is the fastest way to notice you are stalling.

You free up attention for listening. If the first thirty seconds are automatic, you can spend them watching the other person instead of assembling your next clause. That matters, because their first reaction is the most informative thing in the meeting and it is easy to miss while composing.

Why is the opening worth rehearsing more than the rest?

Because it is the only part you control, and it sets the shape of everything after.

The middle of a difficult conversation cannot be scripted — it depends entirely on what the other person says, and a manager working through a memorised sequence while someone is reacting is worse than one improvising attentively. Trying to rehearse it produces stiffness.

The opening is different. You know exactly what you are going to say, nobody has interrupted yet, and it establishes whether this is a direct conversation or an evasive one. It is also the part most damaged by nerves.

Rehearse two things: the first three sentences, and the closing commitment. The structure of both is set out in the complete guide to difficult conversations for new managers.

Does this have anything to do with deliberate practice?

Partly, and it is worth being careful about the claim.

The idea from the deliberate-practice literature that transfers cleanly is this: improvement comes from repeated attempts at a specific skill with feedback, not from accumulated experience. That is why a manager can have twenty years of difficult conversations and still open every one of them badly. Nobody ever told them, so there was nothing to correct.

What does not transfer cleanly is the volume. Deliberate practice in music or sport involves thousands of repetitions, and no manager is going to rehearse a promotion conversation two hundred times. The realistic version is smaller: a handful of run-throughs, on the parts that fail, with something that tells you how it landed.

The feedback half is the part usually missing. Rehearsing alone fixes fluency and cannot tell you that your opening sounds accusatory, because you already know what you meant.

What should you rehearse besides the words?

The pause, and the reaction you are least prepared for.

The pause after the message is where most managers undo their own work, and it is rehearsable in the plainest way: say your opening, then stay silent while counting to ten. It feels absurd alone in a room. It feels considerably worse in the meeting, which is the point of doing it once beforehand.

For the reaction, pick the one you are dreading. If it is anger, say your opening and then say the angriest plausible response out loud, and answer it. If it is tears, or flat denial, do the same. You are not predicting what will happen; you are making sure the first time you hear that response is not in the room. The specific replies to defensiveness — the most common of the three — are in what to do when someone gets defensive.

Why does the same manager keep making the same mistake?

Because nobody tells them, and the conversation itself gives no usable signal.

A conversation that went badly and one that went well can feel almost identical from the manager's side. The person nodded, agreed, said they understood. Three months later one of them has changed their behaviour and the other has quietly started looking for another job, and by then the connection to a specific sentence in a specific meeting is gone.

So the habit persists. The manager who buries the message under two minutes of preamble does it in every conversation for a decade, and experiences it as being considerate. There is no correction because there is no feedback.

This is the argument for practising somewhere the feedback exists. Recording yourself and listening back is crude and works — most people identify their own tell within one playback, and it is almost always the same tell every time. Rehearsing with a peer who will tell you honestly is better. Anything that gives you a response to your actual words, rather than to your intentions, breaks the loop.

The specific thing to listen for: how long it takes you to say the difficult sentence. Most people are surprised.

When is rehearsal not worth it?

For conversations that should happen immediately.

Small course-corrections lose most of their value with delay. If someone talked over a colleague in a meeting this morning, the useful version of that feedback happens today, imperfectly phrased. Booking time to rehearse it converts a thirty-second remark into an event, and an event is a much bigger thing to receive.

The rough line: rehearse anything with a consequence attached or a reaction you are worried about. Do not rehearse anything you could say in the corridor in under a minute.

What does five minutes of rehearsal actually look like?

Say the opening out loud. Notice which sentence came out badly. Fix that sentence. Say it again. Say the closing commitment. Stop.

That is the whole thing, and it is more preparation than most managers do for a conversation that might determine whether someone stays. The version people usually do instead — running it in their head on the way to work — has the same feeling of preparedness and does almost none of the same work.

If the conversation is one of the recurring hard ones, the individual playbooks are worth reading the week you need them: the underperformance conversation and how to tell someone they are not getting promoted are the two that most reward being said out loud first, because both contain a sentence people reliably soften into meaninglessness.