Level 16 min
Open
Nia disagrees with the rating once, then listens. She mentions the inherited accounts as soon as you ask about July.
- One objection
- Shares the cause when asked
Leadership roleplay scenarioPerformance review role play: the underperformer who pushes back
A strong account lead had a weak year and disputes her rating. Practice stating it plainly, separating bad luck from what she could control and agreeing on a plan with Nia, an AI employee who arrives with a list of reasons.
NiaAI account lead
Mood · AI character
I saw the rating in the system. “Partially meets,” after the year I had?
Yes, and I want to show you how I got there. It comes from the three goals we set in January.
Two of those goals were gone the day our biggest account got acquired.
Scored live
- Names the rating early
- Ties it to the agreed goals
- Separates bad luck from choices
- Listens before defending
- Agrees on a plan she owns
AI coachYou named the rating and its basis up front. Now take the goals one at a time.
The situation
A strong performer had a weak year. Now she has to hear it.
Context
You lead the account management team at a B2B software company. Nia has been one of your best account leads for three years and was rated “Exceeds” last year. In January you agreed on three goals with her: grow her book by 10%, renew every account due this year and run a quarterly review with each of her eight largest accounts. Her rating, calibrated with HR last week, is “Partially meets expectations.” She has already seen it in the system.
What you know
- Her book grew 2% against the 10% goal.
- The one renewal she lost was her largest account, after a new owner moved it to a single vendor.
- Her quarterly account reviews stopped in July: three of eight have happened since.
- The rating is final. Her comments go into the review record.
Your goal
- State the rating and the reason early, without apologizing for it.
- Separate what was outside her control from what was in it.
- Agree on a plan for the next six months, with dates she accepts.
The AI character
Meet Nia. She thinks the rating ignores her year.
- Her role
- Account lead for mid-size customers. She owns renewals and growth for her book and has run it for three years.
- Her year
- Lost her largest account in March when its new owner consolidated vendors. Took over four more accounts in July when a colleague left.
- What she cares about
- A rating that reflects what she could control
- Her bonus, which follows the rating
- Being seen as a strong performer again
- What she won’t say first
- She dropped the account reviews in July to keep the inherited accounts from leaving.
- She never told you she was overloaded. She assumed you knew.
- She is afraid the rating means she is being managed out.
- How she reacts
- Vague words like “ownership”She asks for examples and argues harder.
- A fair point concededShe drops the defense and talks about July.
- The rating softened under pressureShe pushes for the next rating up.
- Starting mood
Defensive. She came with a list of reasons and expects the rating to stand whatever she says.
What is scored
Five criteria. Each scored from the transcript.
After each review with Nia, the AI rates what you said on these five criteria and quotes the lines that helped or hurt each score. Your own review form and competency model can replace them.
- 01
Names the rating early
What good looks likeSays the rating and what it is based on in the first two minutes, then explains.
Common missTen minutes of praise first, so the rating lands as an ambush.
- 02
Ties it to the agreed goals
What good looks likeWalks through the three goals set in January and what happened to each one.
Common miss“You didn’t show enough ownership this year.”
- 03
Separates bad luck from choices
What good looks likeSays out loud that the lost renewal was outside her control, then stays on what was in it: the reviews that stopped.
Common missDismisses the acquisition, or lets it explain the whole year.
- 04
Listens before defending
What good looks likeAsks what she wants weighed and sums it up in her words before answering. Empathy is scored here, from what you say.
Common missMeets each objection with a counterargument.
- 05
Agrees on a plan she owns
What good looks likeEnds with two or three goals she helped set, each with a date, and a regular check-in.
Common miss“Let’s touch base next quarter.”
How the conversation can go
From “that rating isn’t fair” to a plan with dates.
Nia · AI account lead
“Partially meets”? I lost our biggest account to an acquisition. That rating isn’t fair.
You
The acquisition is part of the story, and I raised it in calibration. The rating comes from the three goals we set in January. Can I walk you through each one and then hear what you think I got wrong?
AI coach · Names the rating earlyYou kept the rating on the table and promised her the floor. No backpedaling.
Nia · AI account lead
Fine. But growth was never going to happen after March.
You
Agreed, that renewal wasn’t yours to save. What I want to understand is the account reviews. You ran them every quarter until June, then three of eight. What happened?
AI coach · Separates bad luck from choicesYou conceded what was outside her control, then moved to what was in it.
Nia · AI account lead
I inherited four accounts in July. Every week was a fire. The reviews felt like a luxury.
You
So you were running two books, and the reviews were the first thing to go. Is that right?
AI coach · Listens before defendingA summary in her words before any answer. This is where empathy is scored: in what you say.
Nia · AI account lead
Yes. And nobody asked how it was going.
You
That’s fair, and part of it is on me. What would let you run all eight reviews next half?
Nia · AI account lead
Move two of the inherited accounts to the new hire. Then I can do all eight by March.
You
Let’s agree on that. The two accounts move by February 1, all eight reviews happen by the end of March, and we check progress in our one-on-ones every other Tuesday.
AI coach · Agrees on a plan she ownsHer idea, your dates, and a check-in rhythm. She leaves with a plan she owns.
Underperformer review · 5 criteria84/ 100
The plan has dates, but next year’s growth goal never came up. Run it again at the escalating level and close on it.
Difficulty
Three versions of Nia. Start open, finish with a formal challenge.
Level 210 min
Defensive
The default version. She blames the acquisition for the whole year and names the inherited accounts only after you concede the part she couldn’t control.
- Disputed rating
- Hidden workload
Level 314 min
Escalating
Nia says she will contest the rating in writing, asks for HR to join and wants to know whether she is being managed out.
- Formal challenge
- Job security question
- Rating under pressure
Related scenarios
Practice the conversations around it.
Kenji · AI backend engineer
ManagersB2B software
Conflict resolution: “She merged my code while I was out”
Two engineers on your team stopped working together after a merged change and a public revert. Hear one side without taking it and agree on a plan.
Difficulty: IntermediateDuration: 10 min
Omar · AI senior claims handler
ManagersInsurance
Giving feedback to an employee
Your most experienced claims handler skipped a file note, and a customer escalated. Name what happened, find the cause and agree on one change.
Difficulty: IntermediateDuration: 8 min
An underperformance review is one of several talks a manager has each review season. With AI practice for performance review conversations, managers rehearse them all, from a disputed rating to a top performer asking what comes next.
Browse all sales, service and leadership scenariosFAQ
About this scenario.
How do you run a performance review with an underperformer?
State the rating and its basis in the first minutes, then go through the goals you agreed on one by one. Acknowledge what was outside the employee’s control, stay on what was in it, listen before you defend and end by agreeing on two or three goals with dates and regular check-ins.
Can we run this performance review role play without AI?
Yes. A colleague reads Nia’s card and plays her, holding back the July workload until the manager earns it. Another plays the manager, and an HR partner or a peer scores. The third criterion is the hardest for an observer to judge: did the manager concede the lost renewal and still stay on the account reviews? The sample dialogue shows one way through without moving the rating.
How does the AI score the manager?
Line by line. Each of the five criteria gets a score from what the manager actually said, with the quotes that justify it. Listening and empathy count when the words show them, for example when the manager sums up Nia’s July in her own terms. The manager’s emotions are not read. A mood exists only for Nia, as part of her character.
Can we use our own rating scale and review form?
Yes. Load your rating scale and review form into the scenario builder, and Nia disputes your own rating label instead of “Partially meets.” Her goals follow your review form, and the criteria come from your competency model.
Review season is coming. Rehearse the hardest one first.
Run the review with Nia during a short demo, or bring the one your managers dread most.