Hey HN crowd.<p>I'm a new manager in a relatively large company (think FAANG) and just went through a training for employee performance training.<p>For each direct report, we are expected to provide a rating (meet expectations, does not meet, exceed etc.), but in addition to that, we also need to "rank" each individual within each rating, with the goal of helping our director do a "calibration" that will basically curve overall ratings at the org/company level (meaning that someone who meets expectations may be pushed down to "does not meet", purely to match some statistics).<p>My knee jerk reaction here was that this was pretty unethical, after all, if someone does meet expectations and the data is there, it should be possible to give them the corresponding rating / comp increase? But then after talking with other manager friends from other companies, it does seem to be something that a lot of companies do to some extent...<p>So here comes my question: what do you guys think of this practice? Is it an acceptable business practice? Should such companies be named and shamed?
Well, given that performance reviews are usually subjective, it’s important to calibrate those within a business unit to reduce this discrepancy. If A tends to be a tough grader and B tends to be an easy grader things must be normalized to provide equal rewards for equal performance. Otherwise B’s people get all the money. This also serves to train in the new managers into the process and approaches to evaluation.<p>However, if it’s the kind of stack ranking that must yield one “fire” and two “must improve” then it’s evil to use. Imagine, somehow, that a group has all legitimate superstars. One gets tossed and at least two leave with the first review cycle. The rest, being observant, probably leave over the next 6-12 months after such a surprise. This turnover would be perfectly reasonable for a website but somewhat fatal for, say, an auto-driving group.
It is likely that your manager's manager is doing the same thing to you.<p>Your team might meet expectations and the data is there to support it, but your team's rating / your rating may be pushed down to "does not meet" to force a statistical curve.<p>How does it make you feel?
With stack ranking, team members compete, not cooperate. They will make other members look bad, with trash talk, not responding to slack and email, ans actual code sabotage. Some team members act like the chariot race in Ben Hur.<p>Remember that EVERY PLAYER in the winning Super Bowl gets a ring. At the Olympics, EVERY team member gets a metal (Ice Hockey, Basketball, Soccer).<p>Some managers will hire someone incompetent just so they can stack rank fire them.
Stack ranking is a terrible way to measure performance. If managers can’t figure out how to objectively measure performance using tools like skill matrices then they are just bad managers.<p>I’ve left companies because they used stack ranking. When I’ve been in a position of authority I’ve crushed any attempt to introduce stack ranking (or “calibration” or any other variant).<p>First line managers should be able to determine if an individual is performing at, below, or above expectations on their own merits. If they can’t you should fire the second line manager for failing to train the first line manager.<p>Jack Welch introduced stack ranking as a way to force managers to terminate poor performers who were being kept around too long. That’s a distinct issue (under empowered or gutless managers) that should (again) be fixed by training and holding managers accountable.
Rank and yank:<p><a href="https://en.wikipedia.org/wiki/Vitality_curve" rel="nofollow">https://en.wikipedia.org/wiki/Vitality_curve</a><p>Amazon is mentioned there, as well as many others.