One question, not a status round.

A 1:1 becomes a status meeting when the half hour fills with tickets. Coaching is one question, then the room to answer it.

AI prepares the edges. You own the middle. The edge is a short list of questions drafted from last week’s work. The middle is the question you choose, and the silence after you ask it.

Why the 1:1 slides into status

Status is easy to run. There is always a list. The person leaves with the same skill they walked in with, and you leave feeling the meeting happened.

A coached 1:1 is smaller. One question that cannot be answered with a ticket number. You cannot find that question in the last five minutes before the call, and you should not let a model pick it for you.

Before (10 minutes)

Give the agent the work, not the verdict.

• Paste what shipped, what stuck, and what they said they wanted. Facts from last week. No score.
• Ask for five questions a manager could ask. Not advice. Not a diagnosis.
• You pick one. Write it down so you do not drift back to the list.
• If a question only makes sense because of a judgment you did not paste, throw it out. The model guessed. That is the line.

During (fully human)

Ask the question. Then wait.

• Do not rescue a pause with a second question.
• Do not open a tool for “what to say next” while they are talking.
• Capture their words, not a summary of what you hoped they meant.
• Close with one next step they own. Name it out loud.

The agent is not in the room. If it starts supplying the conversation, you have handed it the 1:1.

After (same day)

• One line on what you heard, in their words.
• One next step, with a date.
• If you want a cleaner note, paste those two lines and ask the model to tighten them. It does not grade the person. It does not add a judgment you did not write.

How to brief the person without handing over the judgment

Two kinds of sentence.

A fact: “Last week they led the release. The rollback took a day. They asked to run the next one.”

A judgment: “They are not ready for a larger scope.”

Facts can go in the brief. The judgment stays with you. You may be right. The model does not get to say it first, and it does not get to store it.

Weekly cadence

1. The day before: ten minutes. Work from last week in, five questions out, one question kept.
2. In the 1:1: ask it. Listen. One next step before you leave.
3. Same day: two lines. What you heard. What they own.
4. Do not build a file of model commentary on the person.

Mistakes to avoid

• Walking in with five questions and asking all of them.
• Letting the model choose “the best” question.
• Pasting a private rating so the questions come out sharper.
• Turning the note into a performance record.
• Using a tool the company has not approved for notes about a person.

Tools

Whichever model the company already trusts. The prep card is a page you keep. The model drafts questions from the work you paste. You still walk in with one question.

The 1:1 prep card is in the free kit

New here? Read who writes these notes, browse all weekly notes, or get the free kit below.