I am never going to do another annual performance review ever fucking again

There's a lot I like about management -- and honestly, evaluating performance for feedback purposes is pretty fun! -- but giving humans a rating on "individual" performance is like sandpaper on my soul

Every time I do it I end up thinking

"So, uh, mostly what we're measuring here is, did we as a management team succeed at giving this person high impact person-shaped problems this year?"

Every
Single
Time

An annual performance review is a better review of *me as the manager* than a review of a report

There are definitely folks who are better at finding them-shaped problems and convincing people that they're high impact than others...

... but is that really the most important quality to prioritize in an engineer?

... and if it is, why pretend we're measuring other things?

Related but distinct from this rant, a piece of career advice from the manager trenches:

Your corporate review system does not solve for "make everyone as valuable to the company as possible."

Your corporate review system probably prioritizes making things easy/convenient for overtaxed middle managers over your career growth.

This isn't middle manager perfidy, it's an emergent property of a system where everyone is doing their best.

If your goal is to maximize title and compensation you're probably going to need to be willing to be a little bit weird and a little bit difficult and a little bit pushy and a little bit self-interested.

And a lot dissatisfied.

Not so much that you alienate people, though, which is the tricky thing.

And when I say "dissatisfied" I don't mean with your comp. I mean, with the world. With your product.

You are not going to maximize your comp and title by doing what you're told and solving the problems that people put in front of you.

Or by copying the existing high level people in your org/industry

And, honestly? There's a ton of luck involved. What people happen to observe, what people happen to value.

This is one reason why jumping orgs is a good move when you want to maximize comp, you get to reset expectations *and* roll the dice on getting one that's going to see you and think "oh yeah that's the stuff."

Your manager is *not going to tell you* if being more Difficult would help you advance, or if leaving would help you advance.

There are going to be people who don't have to self-advocate, who just got... made Principal Engineers.

Copying them is not going to help you, they were in the right place at the right time for reasons that probably no longer exist.

Your system is also probably solving for "make the people who designed it feel like they are Competent Managers who are Doing Their Best."

This may or may not correlate with "make the company successful."

Like if I designed a perf review system it would give a lot of weight to "Does this person make young women of color feel safe and powerful? Do young women of color credit them as mentors?"

May not even be correlated with the thing I really care about. But it's observable.

The most salient thing I can say to engineers about their performance review system is

Performance management has feedback cycles measured in *decades.*

The system has the properties you would expect from having to wait this long to find out if it "works."