Why this is hard to get right
Imagine you're the VP of People at a 280-person SaaS company. The CEO has scheduled a two-hour talent calibration session with the full executive team in six days. You need to send a pre-read memo 48 hours before the meeting.
The last time you ran this process, three things went wrong. The memo was 14 pages long, executives didn't read it. The rating definitions were inconsistent across functions, so the Engineering VP and GTM VP argued for 40 minutes about what "high potential" actually means. And no one left with clear accountability for the three people who needed performance improvement plans.
This time, you need a memo that's shorter, sharper, and structured around decisions - not descriptions.
You open your AI assistant and type: "Write a talent review memo for our leadership team meeting next week about our managers."
What comes back is a 600-word HR boilerplate document with section headers like "Performance Overview" and "Development Opportunities." It reads like it was written for a company of 5,000. It doesn't mention the 9-box framework your team uses. It doesn't address succession gaps. And it's definitely not going to help the CEO and CFO align on who gets promoted this cycle.
The problem isn't the AI - it's the prompt.
Without knowing your rating framework, your audience's decision-making style, the scope of roles under review, or the specific outcomes you need from the session, the AI defaults to generic. It produces the document equivalent of a shrug.
When you give the model the right inputs - company stage, scope, framework, audience, and desired decisions - the output becomes something you'd actually send to your CEO. The memo is tight, structured around the 9-box, and organized so each section drives a specific action.
That's the difference between a prompt that frustrates you and one that saves you four hours of writing on a Sunday night.
Common mistakes to avoid
Skipping the Rating Framework Reference
If you don't name your talent rating system (9-box, OKR-based, simple high/mid/low), the AI invents its own. The output won't match your company's language, and executives will spend calibration time debating definitions instead of making decisions.
Omitting the Decision Objectives
A talent review memo exists to drive specific decisions: promotions, PIPs, retention investments, succession slots. Without listing those decisions in the prompt, the AI writes a descriptive summary - useful for documentation, useless for alignment.
Not Specifying the Audience
A memo for a CEO and CFO reads very differently from one for HR leadership. Leaving out the audience causes the AI to default to generic people-ops language that loses executive readers in the first paragraph.
Requesting Too Broad a Scope
Asking for a memo covering 'all managers' across 12 functions produces a bloated document no one reads. Narrow the scope to a specific layer and function set so the memo stays actionable and scannable.
Ignoring Tone and Format Instructions
Talent review memos live or die by their readability in a fast-moving executive meeting. Without explicit instructions to avoid HR jargon and keep sections scannable, the AI defaults to dense, bureaucratic prose that gets skimmed or skipped.
The transformation
Write a talent review memo for our leadership team meeting next week about our managers.
**You are a Chief People Officer drafting an executive talent review and calibration memo.** **Context:** - Company stage: Series C SaaS, ~300 employees - Scope: 18 Director-level managers across Engineering, Product, and GTM - Rating framework: 3x3 performance-potential matrix (9-box) - Calibration session audience: CEO, CFO, and 4 functional VPs **Produce a structured memo that includes:** 1. Purpose and decision objectives for the calibration session 2. Summary of the talent pool distribution across the 9-box grid 3. Top-box talent (3-5 individuals): criteria, retention risk, development plan 4. Bottom-box talent (2-3 individuals): criteria, action timelines, manager accountability 5. Succession pipeline gaps for critical roles 6. Recommended actions requiring cross-functional VP alignment **Tone:** Candid, objective, and executive-ready. Avoid HR jargon. Each section should be scannable in under 60 seconds.
Why this works
Framework Anchoring
Naming the 9-box matrix immediately aligns the AI's structure to a tool your executives already understand. Every section becomes a translation of that framework into narrative, reducing calibration friction during the session itself.
Audience Precision
Specifying the exact audience - CEO, CFO, and functional VPs - shapes both the vocabulary and the political awareness of the memo. The AI learns to write for decision-makers, not HR practitioners, producing a different register entirely.
Decision Orientation
Listing required outputs (top-box retention plans, bottom-box action timelines, succession gaps) forces the memo to be prescriptive. The AI structures each section around a recommended action, not just an observation.
Scope Specificity
Naming the exact number of people under review (18 Directors) and the three functions prevents the AI from writing at the wrong altitude. The output stays appropriately focused for a two-hour executive session.
Tone Guardrails
The explicit instruction to avoid HR jargon and keep sections scannable in under 60 seconds directly counteracts the AI's tendency to default to verbose, technical people-ops language that loses executive readers immediately.
The framework behind the prompt
Talent calibration draws from two intersecting bodies of management theory: performance management research and organizational decision science.
The 9-box grid - widely attributed to McKinsey's work with GE in the 1970s - maps employees on two axes (performance and potential) to create a visual talent portfolio. Research consistently shows that calibration sessions without a shared framework produce wildly inconsistent ratings across functions. One study by CEB (now Gartner) found that 56% of an employee's performance rating is a reflection of their manager's rating tendencies, not their actual performance - what researchers call "idiosyncratic rater effect."
This is why calibration - not just review - is the critical word. The memo's job is to create the conditions for cross-functional alignment before anyone enters the room.
Effective calibration memos apply principles from decision facilitation theory: pre-framing decisions before the meeting reduces in-session anchoring bias and shortens time-to-alignment. They also borrow from structured communication frameworks like SCQA (Situation, Complication, Question, Answer), which ensures every section leads with context, surfaces the tension, and resolves into a recommendation.
When your AI prompt captures these elements - framework definition, decision pre-framing, and audience-aware tone - the output reflects the same principles that make calibration sessions actually work.
Prompt variations
You are a VP of Talent Management at a 2,500-person enterprise preparing for the annual executive talent review.
Context:
- Scope: 42 Senior Managers and Directors across three business units
- Framework: 5-level performance rating combined with succession readiness (Ready Now, Ready in 1-2 Years, Not in Pipeline)
- Audience: CHRO, 3 Division Presidents, and the Board Compensation Committee
Produce a calibration memo that includes:
- Session objectives and ground rules for calibration consistency
- Talent distribution summary by business unit
- Succession-ready talent for 6 critical senior roles
- Flight risk analysis for top-rated individuals
- Equity and representation observations across the talent pool
- Recommended actions with owner and 30-day deadline
Tone: Formal, board-ready, and legally defensible. Avoid subjective characterizations.
You are a founder-CEO conducting your company's first structured talent review with 3 department heads.
Context:
- Company: 65-person Series A startup, 9 managers total
- Framework: Simple high/solid/low performance + high/medium/low potential (informal 2x2)
- Audience: Co-founder and 3 department heads who also manage people being reviewed
Create a short calibration pre-read memo that includes:
- Why we're doing this and what decisions we need to make
- How to use the 2x2 framework consistently
- The 2-3 people we need to align on most (placeholder format)
- One promotion or role expansion to discuss
- Ground rules for the conversation
Tone: Direct, honest, and founder-voice. Short paragraphs. No corporate HR language.
You are an HR Business Partner drafting a mid-year talent calibration memo for a single function (Marketing, 22 people, 6 managers).
Context:
- This is not an annual review - it's a mid-year recalibration to adjust development plans
- Framework: Three-level rating (Exceeding, Meeting, Below) + one flag for flight risk
- Audience: CMO and two senior directors who manage the 6 managers
Produce a focused memo that includes:
- What has changed since January calibration (2-3 key shifts)
- Managers whose ratings have moved significantly and why
- Individuals flagged as flight risk with recommended retention actions
- Development investments to prioritize in H2
- One question the CMO needs to decide before the session ends
Tone: Conversational, data-informed, and action-focused. Keep total length under 600 words.
When to use this prompt
Chief People Officers
CPOs preparing calibration materials for annual or mid-year talent reviews need a memo that drives alignment across executive teams, not just documents scores.
CEOs at Growth-Stage Companies
Founders and CEOs running lean people functions use this to structure their first formal talent calibration process as headcount crosses 100-500 employees.
VPs of Engineering or Product
Functional leaders entering an executive calibration session use this memo to present their team's talent landscape clearly and advocate for high-potential individuals.
HR Business Partners
HRBPs supporting executive teams use this prompt to draft the pre-read memo that sets the agenda and framing for a multi-hour calibration session.
Talent Management Consultants
External advisors helping leadership teams stand up a talent review process for the first time use this to create a template the client can own and iterate on.
Pro tips
- 1
Specify your rating framework by name - whether it's a 9-box, OKR-based rating, or a custom rubric. The AI structures the entire memo differently depending on how performance and potential are defined.
- 2
Include the decisions that must come out of the session. Listing 3-4 specific outcomes (e.g., promote, PIP, retention bonus, succession slot) focuses the memo on action, not observation.
- 3
Name the most sensitive dynamics you're navigating, such as a high performer at risk of leaving or a VP who over-rates their team. The AI can frame those sections with appropriate nuance.
- 4
Adjust the tone instruction based on your company culture. A startup with a direct feedback culture needs different language than a regulated enterprise where every word carries legal weight.
The 9-box grid (performance vs. potential) is the most widely used talent calibration tool in executive settings. When you use it in your prompt, structure the memo output around three zones:
Top-right box (High Performance / High Potential): This is your succession pipeline. The memo should name the 2-4 people in this box, their readiness for the next level, and any retention risks. Executives need to leave the session with a clear investment plan for these individuals.
Middle boxes (Solid contributors): This is the largest group. The memo doesn't need individual deep-dives here. A summary statement about development priorities and manager quality is sufficient.
Bottom-left box (Low Performance / Low Potential): This is the hardest conversation. The memo should name the 1-3 people in this box, the action plan (PIP, role change, or exit), the timeline, and who owns the conversation.
Calibration consistency tip: Include a definitions section at the top of the memo. One paragraph explaining what 'high potential' means in your company's context saves 30 minutes of argument in the session itself. Ask your AI prompt to include this - it's one of the highest-leverage additions you can make.
Some talent calibration memos require extra care. Here are three situations where you need to adjust your prompt:
1. Managing a split rating (high performer, low potential): This is politically charged. The person has delivered results but isn't promotable. Add a specific instruction: 'Include a section for high-performance/limited-potential individuals with language that separates role contribution from career trajectory. Keep framing constructive.'
2. Calibrating across functions with inconsistent standards: When Engineering and GTM use the same rating scale but different standards, your memo needs to surface that inconsistency. Add: 'Note any functional areas where rating distributions suggest calibration drift. Flag for discussion without assigning blame.'
3. Documenting decisions for legal defensibility: If your memo will be used to support terminations or demotions, add: 'Use objective, behavior-based language throughout. Avoid characterizations based on personality or style. Each rating should reference observable actions or results.'
These additions are small in the prompt but produce significant differences in the output's usability and safety.
Before you send your AI-generated talent calibration memo to executive stakeholders, run through this checklist:
Content accuracy:
- [ ] Rating distribution percentages reflect your actual population, not AI-generated estimates
- [ ] Succession gaps match your current open or at-risk roles
- [ ] Action owners are named and have been informed they will be named
Tone and language:
- [ ] No protected-class language anywhere in the document (age, gender, race, disability)
- [ ] All performance characterizations are behavior-based, not personality-based
- [ ] Language is consistent with how your company talks about talent internally
Structural review:
- [ ] Each section has a clear decision or action attached to it
- [ ] Total length is under 2 pages for executive audiences
- [ ] A section exists for 'questions to resolve in session' so the meeting has a clear agenda
Before distributing:
- [ ] HR Legal has reviewed any language related to performance improvement or exit
- [ ] The CHRO or senior HR leader has approved the framing for bottom-box individuals
- [ ] Distribution list is limited to people who need to be in the calibration session
When not to use this prompt
Don't use this prompt for real-time calibration facilitation or live session notes - a memo is a pre-read document, not a meeting tool. It's also not the right format for individual performance reviews or one-on-one feedback conversations, which require a different structure and tone entirely.
If your company hasn't yet defined a talent rating framework, build that first. A calibration memo written around an undefined system will create more confusion than clarity. And if legal or compliance constraints require a specific format, layer those requirements on top of this prompt rather than relying on AI output alone.
Troubleshooting
The AI output reads like a generic HR policy document, not an executive memo
Add two constraints to your prompt: (1) 'Write in active voice with direct, declarative sentences' and (2) 'Avoid HR jargon - use plain business language.' Also specify your audience by title so the AI calibrates register. Generic output usually means the audience and tone instructions are missing.
The memo covers everything but doesn't recommend any actions
Add a numbered list of required decisions to your prompt. For example: '1. Who receives a promotion recommendation this cycle? 2. Who enters a 60-day PIP? 3. Which role needs an external successor identified?' Framing outputs as decisions forces the AI to move from descriptive to prescriptive.
The succession section is too vague and doesn't map to real roles
Name your critical roles explicitly in the prompt - for example, 'succession gaps exist for VP of Engineering and Head of Revenue Operations.' The AI cannot infer which roles are strategically critical without this input. The more specific your role list, the more useful the succession analysis becomes.
How to measure success
A strong AI output from this prompt produces a memo that an executive can read in under 8 minutes and immediately answer: Who are our top 3 people in this group? Who needs a performance conversation? Where are our succession gaps?
Look for these quality signals: each section ends with a recommended action or question, not just an observation. Rating language is behavior-based, not personality-based. The tone is consistent throughout - no sudden shifts into generic HR prose. The succession section maps to real roles, not abstract "critical positions." And the total length stays under two pages, proving the AI prioritized decisions over comprehensiveness.
Now try it on something of your own
Reading about the framework is one thing. Watching it sharpen your own prompt is another — takes 90 seconds, no signup.
a talent review calibration memo for your executive team
Try one of these
Frequently asked questions
Absolutely. Swap in whatever framework your company uses - a simple high/medium/low rating, an OKR-based performance system, or a custom rubric. The key is naming your framework explicitly so the AI structures the memo around the language your executives already know.
Add a note in your prompt that certain individuals require careful framing - for example, a long-tenured employee on a performance plan or a high performer who is a flight risk. The AI can draft those sections with nuance when you flag the sensitivity upfront.
That depends on your calibration process. Some companies use names in pre-read memos; others use role titles or coded identifiers for confidentiality. Specify your preference in the prompt. If you use placeholder names, the AI will generate a realistic template you can populate.
For an executive audience, aim for 1-2 pages maximum. Your prompt should include a word or page count constraint. A two-hour calibration session with busy executives works best when the pre-read is scannable in under 10 minutes.
Yes. Adjust the audience field to 'Board Compensation Committee and CEO,' raise the scope to C-suite and VP-level roles, and add a section for executive compensation implications. The tone should shift to formal and board-ready, which you can specify directly in the prompt.