Best AI Prompts for Reviews

AI summary

Seven AI prompts for review season: self-review drafting, manager reviews of direct reports, peer reviews, calibration briefs, promotion cases, tough feedback conversations, compensation conversation prep.

Prompt 1

Self-Review Drafter

Most self-reviews undersell. This prompt drafts the version that surfaces your actual impact.

Self-review context:

ROLE / LEVEL / TENURE: [BRIEF]
REVIEW PERIOD: [TIMEFRAME]
KEY PROJECTS this period: [LIST]
IMPACT each project produced (data where possible): [BRIEF]
WHAT I LEARNED that made me better: [BRIEF]
WHERE I STRUGGLED: [BRIEF]
MY MANAGER'S STATED PRIORITIES for me this period: [BRIEF]

Draft a self-review:

1. THE OPENING IMPACT STATEMENT: my single most important outcome this period.
2. THE PROJECT-BY-PROJECT impact summaries.
3. THE METRICS where I have them, plainly.
4. WHAT I LEARNED that changed how I work.
5. WHERE I STRUGGLED, surfaced without burying.
6. WHAT I AM ASKING FOR next period.
7. THE QUESTION I want my manager to address.

Do not undersell. Do not over-claim. Surface impact in language a senior leader can repeat.

When to use: Two weeks before the review cycle deadline. · Best model: Claude.

Prompt 2

Manager Draft of a Direct Report Review

Writing reports’ reviews under deadline produces generic language. This prompt structures from your observations.

Report review context (de-identified):

ROLE / TENURE: [BRIEF]
MY OBSERVATIONS this period: [PASTE notes]
KEY MOMENTS that mattered: [BRIEF]
WHERE THEY GREW: [SPECIFIC]
WHERE THEY STRUGGLED: [SPECIFIC]
PEER FEEDBACK if collected: [BRIEF]
THE COMPENSATION CONTEXT and constraints: [BRIEF]

Draft a review:

1. STRENGTHS: 3 anchored to specific observations.
2. GROWTH AREAS: 2 framed as opportunities, with suggested action.
3. IMPACT: paragraph on what mattered most.
4. THE QUESTION I want to explore in the conversation.
5. THE COMP RECOMMENDATION with reasoning.
6. THE TRAJECTORY: what success looks like next period.
7. THE COACHING focus for next period.

Do not invent observations. Build from my notes. Tone: direct, fair, growth-oriented.

When to use: Three weeks before the review meeting. · Best model: Claude.

Prompt 3

Peer Review Drafter

Peer reviews get rushed and generic. This prompt drafts useful feedback from observations.

Peer review (de-identified):

PEER'S role and how I work with them: [BRIEF]
WHAT THEY HAVE DONE WELL: [SPECIFIC moments]
WHERE THEY HAVE STRUGGLED in our work together: [SPECIFIC]
WHAT I HAVE LEARNED from them: [BRIEF]
MY RELATIONSHIP with this peer: [BRIEF]
WHAT FEEDBACK I think they specifically need: [BRIEF]

Draft a peer review:

1. STRENGTHS anchored to specific moments.
2. GROWTH AREA: one concrete area, with the smallest helpful nudge.
3. WHAT I HAVE LEARNED from them.
4. THE QUESTION that would help them grow.
5. THE RECOMMENDATION for impact / promotion if relevant.
6. WHAT I HOPE WE WORK ON together next period.

Do not soften past the point of usefulness. Peer review is most valuable when direct.

When to use: Within review cycle window. · Best model: Claude.

Prompt 4

Calibration Brief

Calibration meetings make or break review accuracy. This prompt prepares the brief.

Calibration context:

MY TEAM and their roles: [BRIEF]
MY PROPOSED RATINGS for each: [LIST]
EVIDENCE supporting each rating: [BRIEF]
MY HARDEST CALL (the one most likely to be challenged): [BRIEF]
WHERE I MIGHT BE BIASED based on recency or relationship: [SELF-AWARE]
THE CROSS-TEAM CONTEXT I should know: [BRIEF]

Draft a calibration brief:

1. EACH RATING with the strongest one-sentence case.
2. THE PUSH-BACK I expect on the hardest call and how to handle.
3. THE BIAS CHECK: where my read might be wrong.
4. THE COMPARISON peers should consider for my proposed top performer.
5. THE CASE for any rating that goes against my prior pattern.
6. THE OPEN QUESTIONS I want calibration to resolve.
7. THE COMMITMENT for follow-up coaching with anyone calibrated down.

Do not let me protect my team beyond what the evidence supports.

When to use: Day before the calibration meeting. · Best model: Claude.

Prompt 5

Promotion Case Drafter

Most promotion cases get written too late and under-evidence. This prompt structures the case.

Promotion context (de-identified):

CANDIDATE'S role and target level: [BRIEF]
THE EVIDENCE supporting promotion: [BULLETS]
WHERE THEY ARE STILL DEVELOPING: [SPECIFIC]
THE EXTERNAL CALIBRATION (what "next level" requires at our company): [BRIEF]
THE COMP IMPLICATIONS: [BRIEF]
WHO WILL READ this case: [BRIEF]

Draft a promotion case:

1. THE RECOMMENDATION stated cleanly.
2. THE EVIDENCE per criterion tied to the next level rubric.
3. WHERE THEY ARE STILL DEVELOPING surfaced.
4. WHY NOW and not next cycle.
5. THE RISK if we wait.
6. THE PROOF that they are already operating at the next level.
7. THE QUESTION the panel will ask: pre-answered.

Do not advocate beyond the evidence. Promotion panels reward calibrated cases.

When to use: 8 weeks before promotion deadline. · Best model: Claude.

Prompt 6

Tough Feedback Conversation

Performance feedback delivered badly destroys trust. This prompt prepares the framing.

Feedback context (de-identified):

REPORT'S situation: [BRIEF]
WHAT THE PERFORMANCE GAP IS: [SPECIFIC, OBSERVABLE]
WHAT I HAVE OBSERVED multiple times: [BRIEF]
WHAT THEY MAY NOT YET KNOW about how this is landing: [BRIEF]
WHAT I WANT THEM TO DO after the conversation: [SPECIFIC]
MY RELATIONSHIP with them: [BRIEF]

Draft a feedback framework:

1. OPENING that does not blindside.
2. THE OBSERVATION stated factually, not characterized.
3. THE IMPACT explained.
4. THE QUESTION inviting their perspective.
5. THE PATH FORWARD co-created.
6. THE NEXT CHECK-IN cadence.
7. THE WARM CLOSE that protects the relationship.

Do not soften past the substance. Tough feedback is care.

When to use: 24 hours before the conversation. · Best model: Claude.

Prompt 7

Compensation Conversation Drafter

Comp conversations go badly when prep is shallow. This prompt structures the prep.

Comp conversation context:

REPORT'S role and performance rating: [BRIEF]
CURRENT COMP and proposed change: [BRIEF]
MARKET DATA for comparable roles: [BRIEF]
INTERNAL EQUITY considerations: [BRIEF]
WHAT I CAN MOVE ON: [BRIEF]
WHAT I CANNOT MOVE ON: [BRIEF]
THEIR LIKELY OBJECTION: [BRIEF]

Draft the prep:

1. THE OPENING: warm, direct, names what we are discussing.
2. THE FACTS: market, performance, internal equity.
3. THE NUMBER OR OFFER plainly.
4. THE REASONING in their language.
5. THE LIKELY OBJECTION and response.
6. THE TRADE-OFFS available.
7. THE NEXT STEPS and timing.

Tone: respectful, grounded, never paternalistic.

When to use: Day before the comp conversation. · Best model: Claude.

Connectors are now standard

Claude, ChatGPT, and Grok all support connectors that let your AI read live data from your work tools (Gmail, Notion, GitHub, Asana, HubSpot, Stripe, and many more) instead of relying on you to paste context. For review writers this means the AI can read your Notion 1:1 notes, your Gmail thread with the report, your calendar.

For performance reviews, the connectors worth pairing with these prompts:

Get Smarter About AI Every Morning

Free daily newsletter. Built for people who want to use AI well, not chase every model.

Free forever. Unsubscribe anytime.

Two ways to go further