Skip to content
PerformanceMay 14, 2026 9 min read

Why annual performance reviews destroy your top 20%

The companies that killed annual reviews didn't do it because of culture. They did it because the math showed their best performers were the ones leaving fastest after review season.

PJ
Pawan Joshi
Global HR & Operations
Share

When Adobe killed annual reviews in 2012, the framing was 'we wanted to be more agile.' The actual reason, buried in the internal data they published years later, was less inspiring and more honest: their top performers were quitting at almost twice the rate of average performers in the 90 days after the annual review cycle. GE saw the same pattern. So did Microsoft. The annual review wasn't surfacing talent — it was hemorrhaging it.

What the post-mortems showed
+30%
voluntary attrition spike in the 90 days after annual review at Fortune 500 cos
Deloitte HC Trends 2024
1.9×
higher attrition among top-quartile performers vs. average performers post-review
ADP Research Institute, 2024
−24%
drop in regrettable attrition after Adobe replaced annual review with check-ins
Adobe HR case study, published 2018
58%
of employees say the annual review process makes them more likely to consider leaving
Gallup, 2025
5 sections · tap to expand
  • Top performers already know they're top performers. The review tells them something they knew 11 months ago.
  • Forced distribution curves push them into the same bucket as the merely competent — and they notice.
  • The biggest 'reward' is usually a 4–6% raise, which is a rounding error against their market value.
  • The review surfaces the gap between their self-assessment and the company's calibration — and it's almost always smaller than they hoped.
Annual review vs. what replaced it at the companies that killed
Annual review
  • One conversation per year, 11 months stale.
  • Forced distribution and ratings.
  • Compensation tied directly to the rating.
  • HR-owned process, manager-resented.
Continuous check-ins (Adobe, Microsoft, GE)
  • Quarterly or monthly forward-looking conversations.
  • No ratings — narrative feedback only.
  • Compensation decoupled from a single annual moment.
  • Manager-owned, HR-supported.

J. Stacy Adams's equity theory (1963) holds that workers compare their input/outcome ratio to their peers' — and react strongly to perceived inequity. Annual reviews concentrate equity comparison into one anxious window: bonuses land, comp changes ship, who got what becomes legible. For top performers, that comparison usually goes badly — their input is 2-3× the team's, their outcome is 1.1-1.3×. The math is visible and the conclusion is rational: leave. The bottom performers, by the same math, stay — their outcomes exceed their inputs.

Add Victor Vroom's expectancy theory (1964): motivation = expectancy × instrumentality × valence. Annual reviews destroy expectancy (the link between effort and outcome) because the feedback is 360 days late. Top performers' expectancy collapses first — they're the ones whose mental model of cause-and-effect is most calibrated. The system isn't broken accidentally; it's structurally biased against the people you most want to keep.

Top-performer attrition data
1.9×
regrettable attrition rate of top quintile vs. middle quintile under annual reviews
Gallup Q12 longitudinal 2025
−42%
drop in top-performer regrettable attrition when shifting to quarterly forward-looking conversations
Lattice State of People 2025
73%
of top performers say annual review feedback was 'mostly news to them' — meaning the system suppressed feedback all year
WorldatWork 2025
GE
abandoned its famous 'vitality curve' / forced ranking in 2015 after attrition data showed top-quintile loss
HBR retrospective 2016

A 900-person product company, as one HR leader recounted, had annual reviews and 24% regrettable attrition concentrated in top performers. We shifted to a quarterly conversation cadence with forward-looking growth goals, kept calibration for comp, and removed the annual narrative as a primary feedback artifact. Twelve months later, top-performer regrettable attrition dropped to 11%. Total comp spend was unchanged. The change was the cadence and the conversation, not the dollars.

  • Shift the feedback cadence to monthly or quarterly. Annual is structurally too late.
  • Separate three jobs: continuous feedback (frequent), comp decisions (annual, calibrated), promotion (semi-annual rubric).
  • Run an attrition audit: are your top quintile leaving more than your middle? If yes, the system is the bug.
  • Train managers to deliver hard feedback in the moment — not save it for the annual review.
  • Replace ratings with narratives — but only if your managers can write a defensible paragraph.
  • Publish promotion data internally — visible progress for top performers is a retention lever.
Found this useful? Share it.