AI for Performance Reviews — Prompts That Save Hours
Performance review season is universally dreaded. Managers spend hours writing reviews that often end up vague and unhelpful. AI can draft specific, actionable reviews in minutes — if you feed it the right information.
The Manager Review Prompt
Prompt:
Write a performance review for an employee in a [role] position. Strengths: [list 2-3 specific accomplishments or behaviors] Areas for growth: [list 1-2 specific areas] Key metrics: [any relevant numbers] Overall rating: [exceeds/meets/below expectations]
Format: 3 paragraphs. First: specific accomplishments and strengths with examples. Second: growth areas framed as development opportunities with actionable suggestions. Third: forward-looking goals for next review period. Tone: constructive, specific, balanced. Avoid vague phrases like “good team player” — use specific examples.
The key instruction: “Avoid vague phrases.” Without this, AI writes generic reviews that could apply to anyone. With it, the output references the specific accomplishments you provided.
The Self-Review Prompt
Help employees write better self-reviews:
Prompt:
Help me write a self-review for my role as [title]. This review period I:
- Accomplished: [list 3-4 things]
- Struggled with: [1-2 honest challenges]
- Goals for next period: [2-3 goals]
Write in first person. Be specific about impact (numbers, outcomes). Frame challenges as learning experiences. Keep each section to 2-3 sentences. Tone: confident but not boastful, honest about growth areas.
Share this template with employees before review season. Better self-reviews lead to better conversations.
Turning Vague Feedback Into Specific Reviews
Managers often have a general sense of performance but struggle to articulate it. AI helps translate:
Prompt:
I need to write a review for someone who is generally a strong performer but needs to improve their communication with other teams. They delivered Project X on time and under budget, but two stakeholders complained about lack of updates during the process. Turn this into a professional performance review paragraph that acknowledges the achievement while addressing the communication gap constructively.
Rating Calibration Language
Prompt:
Write performance review language for each rating level for the competency “[competency name, e.g., collaboration]”:
- Exceeds expectations
- Meets expectations
- Below expectations
Each should be 2-3 sentences with specific, observable behaviors. These will be used as examples for managers writing reviews.
This is useful for HR teams creating review guidelines. Generate calibration language for each competency in your review framework.
Development Goal Setting
Prompt:
Based on this performance review feedback: [paste the growth areas section], suggest 3 SMART development goals for the next review period. Each goal should include: specific action, measurable outcome, realistic timeline, and resources needed (training, mentorship, stretch assignments).
Common Mistakes AI Helps Avoid
Recency bias: AI doesn’t forget Q1 accomplishments when writing in Q4. Feed it the full year’s highlights and it weighs them equally.
Vague language: “Good communicator” means nothing. AI prompted correctly writes “Presented quarterly results to the executive team three times, receiving positive feedback on clarity and data visualization.”
Sandwich feedback: The old “positive-negative-positive” structure feels manipulative. AI can write reviews that integrate strengths and growth areas naturally rather than artificially sandwiching them.
Privacy Considerations
- Don’t paste employee names or identifying information into consumer AI tools
- Use “Employee A” or role descriptions instead
- If your company has an enterprise AI tool (Microsoft Copilot, etc.), use that for employee-specific content
- Keep AI-generated drafts in your secure HR system, not in chat histories
The Review Season Workflow
- Collect inputs: Gather self-reviews, peer feedback, and your notes (this is the human work)
- Draft with AI: Use the manager review prompt with specific inputs (3 min per review)
- Personalize: Add context only you know — relationship dynamics, career aspirations, informal contributions (2 min)
- Calibrate: Compare AI drafts across your team for consistency
- Deliver: The conversation matters more than the document
For a team of 10, this cuts review writing from 8-10 hours to 2-3 hours. That’s time you can spend on the actual conversations.