Skip to content
Playbook
IntermediatePillarHRManagerFounder

Performance Reviews People Don’t Dread

The system around the review matters more than the review itself. A modern approach to goals, feedback, calibration, and the conversation.

12 min read
On this page
60-Second Summary
  • Annual reviews work when they document, not when they discover.
  • Calibration sessions matter more than the rating scale you choose.
  • Tie reviews to growth and comp explicitly — pretending they're 'just for development' fools no one.
  • Continuous feedback + lightweight annual sync is the modern default.

Reviews fail because they’re the one moment in the year a manager catches up on feedback. The fix is upstream — a continuous cadence, then a review that summarizes what both sides already know.

Why reviews fail

  • Feedback only arrives at review time
  • Goals are vague or unmeasurable
  • Ratings are negotiated, not calibrated
  • Compensation is decided separately, with no logic the employee can see
  • Manager spends 6 hours writing what should have been said in 6 conversations

The system around the review

Year-round system
  1. Goals
    Set quarterly
  2. 1:1s
    Weekly feedback
  3. Mid-year
    Lightweight check-in
  4. Self-review
    1 week before
  5. Peer feedback
    Structured prompts
  6. Manager write-up
    Synthesis, not surprise
  7. Calibration
    Cross-team alignment
  8. Conversation
    Two-way, with growth plan

Ratings or no ratings

Trade-offs
Ratings (1–5 or similar)
  • Easier comp linkage
  • Forces a stance
  • Risk of being reductive
  • Calibration overhead
No ratings (narrative)
  • Encourages real conversation
  • Harder to link to comp
  • Risk of avoiding tough calls
  • Needs disciplined writing

Calibration

A meeting where peer managers review proposed ratings to surface inconsistency. Best practice: small groups (5–8 managers), facts before opinions, document the rationale, then close.

The conversation

  1. Lead with the summary, not the prose
  2. Use evidence, not adjectives
  3. Make growth the second half, not an afterthought
  4. Separate the comp conversation by at least a week
  5. End with one written growth focus for next cycle

Frequently asked questions

Should we kill performance ratings?

Deloitte, Adobe, GE, and Microsoft all moved away from forced ratings — but most reverted to some form of categorical assessment within 3–5 years because comp decisions still need a defensible distribution. The modern consensus: keep lightweight ratings, decouple them from a heavyweight annual ritual, calibrate harder.

Annual, semi-annual, or continuous?

Continuous feedback for development; semi-annual for documented assessment; annual only for comp and promotion decisions. Trying to make a single ritual serve all three purposes is why most review systems are universally hated.

How long should a written review be?

One page of evidence + one page of forward-looking development. Anything longer reads as box-ticking; anything shorter lacks the specificity a calibration session needs.

Can we run reviews without a tool?

Up to about 50 employees, yes — a shared template works. Above that, you need a system of record for legal defensibility and calibration analytics. Lattice, 15Five, and Culture Amp are the dominant choices for 50–2000 employees.

Written by Pawan Joshi.Sources cited inline.
First published 21 Nov 2025See site changelog →