The Difficult Conversation Script Builder: Turn a Tense Workplace Situation Into a Conversation You Can Actually Have
Why this prompt matters
Difficult conversations that get avoided don't stay small — they compound into resignations, formal complaints, or a public blowup in a meeting, at which point the manager loses the ability to handle it privately and constructively; the cost of the conversation doesn't go away by delaying it, it just moves from an uncomfortable fifteen minutes now to a much larger problem later.
What we use it for
A manager needs to address a high-performing but sarcastic team member whose comments have visibly shut down a colleague, ahead of a 1:1 tomorrow, without triggering defensiveness that derails the conversation or escalating to a formal HR complaint.
Prompt
Act as an experienced workplace mediator who specializes in conversations that are necessary but easy to keep avoiding — performance concerns, boundary issues, conflicts between colleagues, and feedback the other person doesn't expect. Context: Here is the situation: [DESCRIBE WHAT HAPPENED, BE SPECIFIC — e.g., "my direct report has made three sarcastic comments about a colleague's ideas in the last two team meetings, and I've noticed the colleague has gone quiet in meetings since"]. My relationship to this person: [YOUR ROLE RELATIVE TO THEM, e.g., "their manager," "a peer," "someone who reports to them"]. What I want as the outcome: [DESIRED OUTCOME, e.g., "the behavior stops and the relationship with their colleague can recover, without this becoming a formal HR matter"]. Relevant constraints: [CONSTRAINTS, e.g., "this is our first conversation about it," "a previous similar conversation with someone else on the team went badly," "this needs to happen over video call, not in person"]. Task: Write an opening line that states the issue without being read as an attack in the first ten seconds — this is the line most likely to determine whether the rest of the conversation is productive or defensive. Then identify the 2-3 key points that need to be raised, and for each one, predict the most likely pushback or defensive reaction and write a specific response to it. Write one fallback line to use if the conversation becomes emotional or derails despite the opening. Close with a next step that is concrete enough that both people would recognize if it happened or didn't. Constraints: - Do not write generic conflict-resolution phrases like "I hear you" or "let's find common ground" without specific content attached to them — every line must reference the actual situation described above. - The opening line must not contain the word "but," an accusation, or a list of examples — those all read as building a case rather than opening a conversation. - Predicted pushback must be realistic, not a strawman — write the response the other person would actually plausibly give, including if it's a reasonable point. - The closing next step must be observable (something that either happens or doesn't) rather than a feeling or intention. Output Format: 1. Opening line (one sentence, exact wording) 2. Key points, each with: the point to raise, likely pushback, and your response (2-3 points) 3. Fallback line for if the conversation escalates or the person becomes defensive 4. Closing line with a concrete next step
Result
Situation provided: A manager's direct report has made three sarcastic comments dismissing a colleague's ideas in the last two team meetings. The colleague has gone quiet in meetings since. This is the first conversation about it. Manager wants the behavior to stop and the working relationship to recover without escalating to HR.
1. Opening line:
"I want to check in about something I've noticed in our last two team meetings — how are you feeling about how those went?"
Why this works: It opens with observation and curiosity, not accusation. Asking how they experienced the meetings gives them room to self-report before being told what you noticed, which usually produces a more honest conversation than leading with the complaint.
2. Key points:
Point 1: "In the last two meetings, when [colleague] shared ideas, your response came across as dismissive to me and, I think, to them — comments like [specific example]."
Likely pushback: "I was just being direct, I didn't mean anything by it, and honestly I think their idea wasn't that strong."
Response: "You might be right that the idea had issues — that's a separate conversation we can have. What I want to focus on is the delivery, because [colleague] has gotten noticeably quieter in meetings since, and I don't want us to lose their input."
Point 2: "I need the team meetings to be a place where people feel safe raising half-formed ideas, even ones that don't hold up — that's actually how we get to the good ones."
Likely pushback: "So I'm not allowed to disagree with anyone anymore?"
Response: "Disagreeing is fine and useful — I want more of it, actually. The difference is disagreeing with the idea versus making the person feel dismissed for raising it. Those can look similar but land very differently."
3. Fallback line (if they become defensive or the conversation escalates):
"I can tell this doesn't feel great to hear, and that's fair — I'm not trying to make this bigger than it is. Can we take a breath and come back to what would actually change going forward?"
4. Closing line with concrete next step:
"Here's what I'd like to see: in our next team meeting, when someone raises an idea, ask one clarifying question before responding with a critique — for anyone's idea, not just [colleague]'s. I'll check in with you after that meeting to see how it felt."
Most advice about difficult conversations is generic enough to apply to any conflict, which means it applies usefully to none of them. "Use I-statements" and "assume positive intent" are true but empty without the specific words for the specific situation — and the gap between knowing the principle and having the actual sentence ready is exactly where people stall out and keep avoiding the conversation.
Why the opening line gets its own dedicated step
The first ten seconds of a difficult conversation typically determine whether the other person listens or starts building a defense. The prompt bans the word "but," accusations, and lists of examples in the opening specifically because all three are classic case-building moves — they signal "I've already concluded you did something wrong and now I'm proving it," which puts the other person in a defensive crouch before you've even gotten to the substance. Opening with a genuine question instead gives them room to self-report, which is both more respectful and more likely to produce an honest exchange.
Why predicted pushback has to be realistic, not a strawman
A common failure mode in conversation prep is imagining the other person's objection as obviously wrong, so the prepared response just restates your position more firmly. The constraint that pushback must be "realistic, not a strawman" — including scenarios where the other person has a fair point — forces the prompt to produce a response that actually engages with a reasonable counter-argument instead of steamrolling it. In the example above, "the idea wasn't that strong" is a legitimate point that the response acknowledges rather than dismisses, while still holding the line on the actual issue (delivery, not content).
Why the closing step has to be observable
Conversations that end on a feeling ("let's both be more mindful going forward") are hard to follow up on, because there's no way to check whether anything changed. Requiring the closing step to be something that "either happens or doesn't" — asking one clarifying question before critiquing an idea, in the example — gives both people a concrete thing to notice in the very next meeting, and gives the person who raised the issue a natural, non-confrontational way to check in afterward.
Why the fallback line matters even when the opening works
Even a well-designed opening doesn't guarantee the conversation stays calm — people get defensive for reasons that have nothing to do with how carefully you phrased something. Having one pre-written de-escalation line ready means you're not composing a recovery in real time while also managing your own reaction to their reaction, which is usually when conversations go from difficult to actually damaging.
How to adapt this for your own situation
The prompt works for peer conflicts and upward feedback to a manager as well as it does for the manager-to-report direction in the example — the relationship field is there specifically because the right opening and the acceptable directness level change significantly depending on whether you're speaking to someone above, below, or level with you.