[UltraLutra]

It’s cute they think managers are evaluated on the quality of their employee performance review.

I don’t disagree with the content. A well-thought review will help employees perform better. AI can’t create a good performance review from whole cloth. You have to at least put some thought into the content and bullet out what you want the review to say. An AI could clean up the language, but not create good content.

But most managers’ managers would consider that wasted effort. The unspoken fact is that performance reviews are pretty arbitrary and end up reflecting whatever is needed at the time. If budgets are tight, performance suddenly becomes less impressive and no one deserves a promotion. If hiring is difficult and an employee expresses dissatisfaction with their pay then suddenly they’re a star performer. The efficient manager tells the AI “make this a good review that makes a case for promotion” or “make this a mediocre review with room for improvement” and lets the AI write something vague and unhelpful that agrees with the result they want. An _effective_ manager will put time into their reviews and growing talent. Managers aren’t incentivized to be effective leaders though.

Source: my experience; your mileage may vary.

[bicx]

I spent a while in management, and what I learned about the review process was pretty eye-opening. Basically everyone was graded on the curve, and you couldn’t afford to have too many “exceeds expectations” ratings. That would mean they probably met expectations above their role, and they could make a strong case for promotion. If you can’t give out lots of promotions, then the top performers get pissed. So, just don’t let them all know how great they’re doing.

Performance reviews are a dumb system, and you shouldn’t be learning anything new from them anyway. Just give helpful feedback throughout the year, and give promotions to people strategically as you’re able.

[WalterBright]

> Performance reviews are a dumb system

The purpose of them is to avoid discrimination lawsuits.

[reval]

When everybody exceeds expectations then the expectations are too low.

[bicx]

I'm talking more about cases when there are (for instance) initially 4 "exceeds expectations" seniors out of a department of 50, and management really just wanted a max of 2. So they curve things more harshly until only 2 get "exceeds expectations."