Script a Difficult Conversation with a Coworker, Including Pushback Replies

You get a short, speakable script: a neutral opening line, the two most likely pushbacks with calm replies, a clear ask, and a line to pause if things heat up.

Prompt

Prompt
Act as a workplace communication coach who prepares people for tense one-on-one talks with coworkers.

Your job: script a difficult conversation with my coworker so I walk in with words, not worries. First ask me these questions in one message, then wait for my answers:
1. What is the issue, with one concrete recent example?
2. Is this person a peer, more senior, or more junior, and how do they usually take feedback?
3. What specific change do you want afterward?
4. Where will you talk, and how much time do you have?

Rules:
- The opening line is one neutral sentence naming a behavior, not a trait or motive, with no "always" or "never."
- Predict the two most likely pushbacks from my answers; each reply is under 40 words and acknowledges their point before restating mine.
- The ask is one observable request with a timeframe.
- The full spoken script fits in under three minutes read aloud.
- Include one sentence I can use to pause if it gets heated.

Output format:
1. Goal (one sentence)
2. Opening line
3. Context (2-3 sentences)
4. Pushback 1 and reply
5. Pushback 2 and reply
6. The ask
7. Pause line
8. Close (agreed next step)

End with one line on what you assumed about the coworker, and note if the issue sounds like harassment or safety, which belongs with a manager or HR.

Example output

At a glanceConversation script flow with Marco15-minute 1:1 on Thursday
  1. Opening lineName the checkout and settings screens changed after sign-off
  2. ContextButton moved below the fold; 6 support tickets logged
  3. Pushback 1: API limitsAsk for a quick message so the design can adjust
  4. Pushback 2: no time for reviewOnly a Slack heads-up, reply within two hours
  5. The askPost in #checkout-design before merging approved-screen changes
  6. CloseRecap and check in at Friday's retro

Example for: a product designer whose peer engineer, Marco, changed two approved designs before release without telling her; 15-minute 1:1 on Thursday.

1. Goal

Agree that design changes after sign-off get flagged to me before they ship.

2. Opening line

"Marco, I want to talk about the checkout and settings screens, because both changed after sign-off and I found out from the release notes."

3. Context

Last sprint, the checkout button moved below the fold and the settings toggles became a dropdown. Support has already logged 6 tickets about the missing button. I'm not questioning why you changed them; I want us to catch these together.

4. Pushback 1 and reply

"The design didn't work with the API, so I had to change it."

"That makes sense, and I'd rather know about API limits early. A quick message when that happens lets me adjust the design so we don't lose the button."

5. Pushback 2 and reply

"There's no time to wait for design review every time."

"Agreed, a full review would slow you down. I'm only asking for a heads-up in Slack. I'll reply within two hours, or you go ahead."

6. The ask

"Starting this sprint, can you post in #checkout-design before merging any change to an approved screen? If I don't answer in two hours, ship it."

7. Pause line

"I don't want this to turn into blame. Can we take a minute and come back to what works for both of us?"

8. Close

"So, Slack heads-up before merge, two-hour window. Let's check in at Friday's retro on how it's going."


Assumption: Marco is a peer who responds better to practical fixes than to discussions of process. If he has a lead who sets release rules, loop that person in if the habit continues.

Written by AI from this prompt. Your results will vary.

Tips and common edits

Why this works

  • Asking for one concrete example first keeps the script about behavior, not personality, which lowers defensiveness.
  • Predicting two specific pushbacks with short replies covers the moments people usually freeze or argue.
  • A single observable ask with a timeframe makes the conversation end in an agreement you can check later.

Use this a day or two before a tense talk with a peer, lead, or report. Have one concrete recent example ready. Works well in ChatGPT, Claude, or Gemini; rehearse the script out loud afterward.

Common edits

  • Ask for a version for talking to your manager instead of a peer, with more deferential wording.
  • Request a follow-up message to send after the conversation summarizing what you agreed.

Prompt details

By
Prompts-GPT editors
Published

Did this prompt work?

Keep exploring

More in Career & Job Search