Team reality

Two Words That Killed Performance Reviews

'Meets expectations.' Nearly every employee got it, because the alternative cost the manager a conversation nobody wanted to have.

Two Words That Killed Performance Reviews
Illustration · Deimar Gutiérrez

42 performance reviews came in one cycle. 39 of them said meets expectations. Two said exceeds. One said below. Twelve managers submitted them independently, none comparing notes, and the distribution was the predictable output of a rubric nobody had to defend.

Meets expectations is what a manager writes when the alternative costs a conversation. Exceeds forces a defense of why this person outran their peers, which surfaces a comparison most managers would rather leave buried. Below triggers a development plan, an HR check-in, and a hard talk about specific gaps. Meets asks for nothing. It's the box a manager ticks when they haven't evaluated the person at all.

So the rubric collapses to one bucket and emits no signal. Compensation calibrates raises against it and learns nothing. The promotion committee ranks candidates against it and learns nothing. The employee reads it and learns nothing. The process still runs, still eats a week of write-up time twice a year, and still produces nothing anyone can use. That's the pattern behind a review cycle where everyone passed: motion without measurement.

Forced distributions get proposed, then killed, because managers hate them. That discomfort is the feature. Tell a manager to rank ten percent of the team as exceeds and ten percent as below, and they have to evaluate the team for real. The ranking produces the conversation meets was dodging.

A lighter version works too: require written justification for every rating, meets included. A manager writing meets now has to name what exceeds would have taken. The friction is small and constant. Faced with defending meets against a specific person, most managers find the honest rating was exceeds or below. The written defense drags the real distribution back into the light the lazy default had hidden. It does one person at a time what good feedback does in conversation.

A review system that survives either method starts producing information. Comp calibrates against real performance. Promotion candidates get compared on real data. The employee finally learns where they stand. Most companies keep the broken version because the broken version is comfortable and the working one is not.