Verified against ChatGPT · 2026-08-11
Write a promotion justification memo that survives a calibration committee's scrutiny
Turns a manager's case for promoting someone into a memo built on specific evidence against the next level's actual bar, structured to hold up when a calibration committee pushes back.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Write a promotion justification memo for a calibration committee that hasn't worked with this person day to day and will push back on anything that reads as manager advocacy rather than evidence.
PERSON AND PROPOSED MOVE
Devon Okafor, Software Engineer L3 to L4.
NEXT LEVEL'S ACTUAL BAR
L4 requires independently owning a subsystem's technical direction and mentoring at least one L2/L3 without direct oversight.
SPECIFIC EVIDENCE
Led the migration of the notifications service off the legacy queue independently over 2 quarters; has been informally mentoring a new L2 hire since March with positive feedback from that hire.
COUNTERARGUMENT YOU EXPECT
Committee may say one migration project isn't enough breadth to prove sustained L4-level ownership.
Open with the promotion ask stated as a claim against the next level's specific bar, not a general "they're ready" statement — "meets the L4 bar because they've independently owned two cross-team technical decisions in the last two quarters" is a checkable claim; "they're clearly operating at the next level" is not. For each piece of evidence given, connect it explicitly to one specific element of the level bar rather than listing accomplishments and leaving the committee to map them itself — the committee's job is to judge the connection, not reconstruct it.
Address the expected pushback directly and by name, in its own section, rather than hoping the evidence section preempts it implicitly. If the honest answer to the pushback is that the evidence is thinner in that specific area, say so plainly and state what would need to be true to fully close that gap — a memo that oversells a weak spot loses credibility faster than one that acknowledges it and argues the rest of the case still clears the bar.
Do not use comparative language against specific named peers ("more ready than X") — anchor every claim to the level bar itself, not to a relative ranking, since a committee evaluates against the bar, not a beauty contest between individuals. Do not inflate a single strong quarter into a pattern if the timeframe given doesn't support that framing — state the actual timeframe the evidence covers.
OUTPUT FORMAT
1. The promotion claim, stated against the specific level bar.
2. Evidence section, each item explicitly mapped to a bar element.
3. A dedicated "anticipated pushback" section addressing the expected counterargument honestly, including any real gap.
4. A one-paragraph recommendation stating whether the case clears the bar as-is or what would need to be true to close it.Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Calibration committees exist specifically to catch cases where a manager's genuine belief in their report substitutes for evidence against a shared bar, so a memo that reads as advocacy rather than argument gets discounted immediately regardless of whether the underlying case is actually strong — requiring every evidence item to be explicitly mapped to a specific bar element forces the memo into the form a committee can actually evaluate quickly, rather than one it has to reverse-engineer, which matters practically because committees review many cases in one sitting and an unmapped list of accomplishments reads as unprocessed input rather than a finished argument. The dedicated pushback section exists because GPT-5.1, asked to write a persuasive memo, will default toward presenting the strongest possible case and omitting weaknesses unless explicitly instructed otherwise — that omission is exactly what makes advocacy memos fail in front of a skeptical committee, since the first question calibration reviewers ask is usually the obvious counterargument, and a memo that has visibly already answered it honestly reads as more credible, not less, than one that pretends no counterargument exists. Prohibiting comparative language against named peers matters for the same reason it matters in the compensation-review prompt: a bar-referenced claim is verifiable against a shared standard, while a relative claim ("more ready than X") just imports a different, unstated bar (be better than a specific person) that the committee never agreed to evaluate against, and also creates an unnecessary and unfair comparison on the record. Explicitly requiring the actual evidence timeframe to be stated protects against a subtler failure mode where one strong quarter gets rhetorically inflated into "a consistent pattern," which a model optimizing for persuasiveness will do by default unless told to be precise about how much time the evidence actually spans.
What you get back
Claim: Devon meets the L4 bar of independently owning a subsystem's technical direction and mentoring without oversight. Evidence: led the notifications-service migration off the legacy queue independently over 2 quarters (maps to independent technical ownership); has informally mentored a new L2 hire since March with positive feedback (maps to mentoring criterion). Anticipated pushback: committee may argue one migration project is too narrow to prove sustained ownership — this is a fair concern; the mentoring evidence and the migration's cross-team scope partially offset it, but a second independent project would close this gap more fully. Recommendation: case clears the bar on balance, though a second data point next quarter would remove the one open question.
Verified against
ChatGPT GPT-5.1 · 2026-08-11
Changelog
- 2026-08-11 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
