TL;DR / The Direct Answer: Giving critical feedback to a defensive employee is one of the hardest parts of being a manager. If you go in unprepared, emotions take over and the message gets lost. Use AI Voice Mode to roleplay the conversation beforehand. Anchor your feedback in the SBI (Situation-Behavior-Impact) model, and have the AI act as a difficult, defensive direct report. You'll get your stumbling blocks out of the way before the real meeting starts.
Who this is for: Managers who need to give critical behavioral feedback.
Skip this if: Managers doing routine status updates.
Note: AI pricing, plan names, and product features can change quickly. Re-check official pages before you pay for a tool or choose a plan.
The Problem: The Dreaded 1-on-1
You have a direct report who is highly skilled, but they've been speaking over junior colleagues in meetings. Or they've been quietly missing deadlines and making excuses. You know you have to address it in your next 1-on-1.
You write down your talking points. You tell yourself to stay calm. But the moment you deliver the feedback, they get defensive. They deflect. They bring up something you did wrong last month. Suddenly, you're off-script, arguing about technicalities, and the core issue remains completely unresolved.
Managers often practice presentations, but they rarely practice the conversations that actually define their culture.
The Framework: Situation, Behavior, Impact (SBI)
To keep feedback objective and de-escalate defensiveness, use the SBI model:
- Situation: Anchor the feedback in a specific time and place (e.g., "In yesterday's sprint planning...").
- Behavior: Describe exactly what happened, objectively, without judging intent (e.g., "...you interrupted Sarah three times while she was presenting her estimates...").
- Impact: Explain the result of that behavior (e.g., "...which caused her to stop sharing, and we missed out on her technical input.").
The model is simple. Executing it while someone is arguing with you is hard. That's where AI comes in.
The RTCE Prompt: The Defensive Roleplay
Use the Voice Mode on ChatGPT or Claude app on your phone, and read this prompt to set up the simulation.
Mistakes to Avoid
- Practicing with text instead of voice: Typing out responses gives you time to think. Real conversations happen in real-time. Use the voice feature on your mobile app so you can feel the pressure of the pause.
- Letting the AI be too easy on you: If the AI accepts your feedback immediately, tell it to "be more resistant and defensive." You want the simulation to be harder than the real thing.
- Forgetting to ask for a critique: After the roleplay, step out of character and ask the AI: "As a management coach, rate my delivery. Did I stick to the SBI model? Where did I get defensive myself?"
The Prompt
Role: You are an employee named [Name] who is highly capable but tends to be defensive when given constructive feedback. I am your manager. Task: We are going to roleplay our 1-on-1 meeting. I will practice giving you feedback using the Situation-Behavior-Impact (SBI) model. You must act defensive, deflect the blame, or make excuses, forcing me to practice de-escalation and staying on track. Context: The feedback I need to give is about [describe the issue, e.g., missing the last two code freeze deadlines]. Example: If I say "You missed the deadline," you might respond with "Well, the requirements weren't clear, and nobody helped me." Wait for me to speak first. Let's begin.
The Common Trap
The common trap is confusing 'being nice' with 'being clear'. Managers often soften critical feedback so much that the employee leaves the room completely unaware they are underperforming. Practicing beforehand ensures you stick to the facts without getting emotionally hijacked.