Performance Review Writing
Parent: Writing and Documentation · Published reference · snapshot 2026-09-08 · skill career-and-formal-writing/references/performance-review-writing.md
↓ Facts as markdownall context files
Performance reviews are not new information — they are synthesis. The cardinal rule is "no surprises": anything that appears in a written review should already have been said in a one-on-one.
These notes link each claim to its source. Some citations identify archived research reports that are no longer hosted on this site. A published reference means the content is available; it does not certify independent review or accuracy.Read the editorial policy and follow the sources before relying on a claim.
Overview
- Performance reviews are not new information - they are synthesis. The cardinal rule is "no surprises": anything that appears in a written review should already have been said in a one-on-one. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#overview]
1. The "no surprises" rule
- If a written performance review contains feedback the employee is hearing for the first time, the manager has failed earlier in the year. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#1-the-no-surprises-rule]
2. SBI (Situation-Behavior-Impact) — Center for Creative Leadership
- Situation - When and where the behavior occurred. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#2-sbi-situation-behavior-impact-center-for-creative-leadership]
- Behavior - What the person actually did. Observable, factual, non-interpretive. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#2-sbi-situation-behavior-impact-center-for-creative-leadership]
- Impact - What followed from the behavior. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#2-sbi-situation-behavior-impact-center-for-creative-leadership]
3. The "behavior + impact + recommendation" structure for review bullets
- Weak: "Sarah is a great team player and is always willing to help." [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#3-the-behavior-impact-recommendation-structure-for-review-bullets]
- Strong: "Sarah led the cross-team migration of the billing service in Q2, partnering with three teams and coordinating four deployment windows; the migration shipped two weeks ahead of plan with zero customer-visible incidents. Next step: I'd like to see Sarah take on a bigger system design role in Q4." [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#3-the-behavior-impact-recommendation-structure-for-review-bullets]
4. Growth-area framing (not "weakness")
- Weakness phrasing (avoid): "Tom is too quiet in meetings." [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#4-growth-area-framing-not-weakness]
- Growth-area phrasing (use): "In larger forums (5+ people), Tom contributes less than I think he could. He has the strongest understanding of the search service on the team, but in the Q3 architecture reviews his ideas often landed in the followup chat rather than in the meeting itself. Growth focus for next cycle: Tom and I have agreed he'll lead two of the next four architecture reviews." [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#4-growth-area-framing-not-weakness]
5. Radical Candor — Kim Scott
- Two axes: how much you care personally about the person, and how directly you challenge them. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#5-radical-candor-kim-scott]
- Radical Candor: Care + Challenge. The target. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#5-radical-candor-kim-scott]
- Ruinous Empathy: Care without challenge. Soft, vague, no real feedback. The most common manager failure mode. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#5-radical-candor-kim-scott]
- Obnoxious Aggression: Challenge without care. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#5-radical-candor-kim-scott]
- The failure mode in writing is almost always Ruinous Empathy - burying real feedback in praise sandwich. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#5-radical-candor-kim-scott]
6. Camille Fournier's patterns
- Lead with strengths, not weaknesses. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#6-camille-fourniers-patterns]
- Use concrete examples or cut the point. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#6-camille-fourniers-patterns]
- Negative feedback goes in private. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#6-camille-fourniers-patterns]
- Most of your time should go to talents and achievements. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#6-camille-fourniers-patterns]
7. Evidence-based claims with concrete examples
- Every assertion should be tied to a specific example. Sweeping claims without examples are the single biggest source of perceived unfairness. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#7-evidence-based-claims-with-concrete-examples]
- Weak: "She has strong communication skills." Strong: "When the platform migration ran into the dependency issue on the search team, she wrote the four-paragraph summary that unblocked the cross-team decision." [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#7-evidence-based-claims-with-concrete-examples]
8. Self-review traps
- False-modesty traps: [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#8-self-review-traps]
- Hedging with "I tried to" and "I worked on" instead of "I led" and "I shipped." [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#8-self-review-traps]
- Crediting the team for outcomes you personally drove. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#8-self-review-traps]
- Listing the work without claiming the result. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#8-self-review-traps]
- The fix: "I led the migration; the team of four shipped it" beats both "I migrated the system" (overclaim) and "We migrated the system" (underclaim). [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#8-self-review-traps]
9. Bias-checking before submission
- Are claims tied to behaviors and outcomes, or to personality? [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#9-bias-checking-before-submission]
- Have I used different language for different people on the same team for the same observed behavior? [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#9-bias-checking-before-submission]
- Have I disproportionately commented on style versus substance? [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#9-bias-checking-before-submission]
Anti-example: the "praise sandwich"
- Don't write: > "Maria is a wonderful teammate and a hard worker. She sometimes can be a bit late on deliverables but overall she's been a great addition to the team." [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#anti-example-the-praise-sandwich]
- The real feedback ("late on deliverables") is buried in vague praise, has no specific example, and gives Maria no idea what to change. [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#anti-example-the-praise-sandwich]
References
- Radical Candor - Our Approach (Kim Scott) [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#references]
- Camille Fournier - The Manager's Path (O'Reilly, 2017) [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#references]
- Center for Creative Leadership - SBI Feedback Model [archived source: https://llms-explorer.com/sources/mdb-context-hub/performance-review-writing/#references]
Where this helps
- Writing a review for a report whose year included a real performance issue that needs to land clearly instead of vanishing into a praise sandwich. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Coaching a manager who defaults to Ruinous Empathy — soft, vague reviews that avoid real challenge — toward Radical Candor's care-plus-challenge target. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Reviewing a self-review draft that hedges every accomplishment and needs to be tightened into ownable, specific claims. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Auditing a batch of reviews across a team for bias before submission — checking whether the same behavior gets described differently depending on who exhibited it. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
How to apply this
- Draft every negative or growth-area point as SBI (Situation-Behavior-Impact) before writing the review prose, so the point stays observable and factual instead of drifting into personality judgment. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Before submitting a review, run the bias-check questions across every report on the team at once, comparing language used for the same behavior across different people, not just proofreading one review in isolation. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Rewrite every sweeping claim by attaching the one specific example that proves it, or cut the claim entirely if no example exists. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Apply the "no surprises" rule as a mid-cycle check, not just at review time — if a point can't be traced back to something already said in a one-on-one, say it in the next one-on-one before it goes in writing. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
Antipatterns
- Burying real, actionable feedback inside a praise sandwich — vague positive statements wrapped around one buried negative point that gives the person no idea what to actually change. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Writing a review point the employee is hearing for the first time — a written review is synthesis of things already said, not a venue for new information. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Labeling a growth area as a fixed trait or "weakness" instead of describing the specific, situational behavior and a concrete next step. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- In a self-review, crediting the team for an outcome you personally drove, or hedging with "I tried to" instead of "I led" — both underclaim and overclaim distort the record. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
Limitations
- SBI describes what happened and its effect, but it doesn't by itself resolve whether the behavior was actually within the person's control or driven by a structural or team problem. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- The "no surprises" rule assumes a manager has been running real one-on-ones with real feedback throughout the cycle — applied to a manager who hasn't done that groundwork, it mainly reveals the manager's own gap. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Bias-checking by comparing language across reports on the same team catches inconsistent phrasing, but it can't catch bias in which behaviors get noticed or documented in the first place. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
- Radical Candor's care-plus-challenge framing describes a target, not a formula — the same direct language that reads as caring challenge from a manager with an established trust relationship can land as pure aggression from one who hasn't built that trust yet. [source] — AI-suggested, synthesized from this pack's existing facts — not extracted from a source document.
Children
- No children recorded.