[lnsru]

At the end of the day the decision is made on personal level. “I like you” or “I don’t like you”. Such frameworks are just the limits how a manager can show liking/disliking someone. The thing is, that measuring someone’s performance is very hard. It’s not standardized multiple choice written exam. And maybe the company does not need more than quick&dirty, but easy to explain solution.

[yodsanklai]

I don't think it's true. It's in the company's interest that people are evaluated fairly, especially if they are going to give big bonus or agressively fire low performers (like some of these tech companies do).

Don't know about Dropbox in particular, but I know other tech companies who take evaluation very seriously. The manager doesn't have the final word and there's a calibration committee that ranks employee based on various data.

It's not easy to measure performance, but not impossible if you put the resource to do so.

I think what is difficult is to align employees incentives with the company's goals. For instance, focus on impact and metrics can back fire.

For instance, in my team, I have an ambitious and hard working colleague who is playing the company's game (basically, checking all the boxes needed go get promoted, playing the evaluation's game). I'm convinced it's not in the company's interest. It builds technical debt, we now have several half-baked time consuming projects that the team can't reasonable handle, dubious metrics, experiments, and systems to produce them that require maintenance etc etc... I'm not blaming him though, that's what the company is optimizing for.

[interactivecode]

Work that makes your boss (or your bosses boss) happy gets you job security and in some companies gets you promoted. If that aligns with “what is best for the company” or an even more distant metric “what is best for the customer“ then thats great, but its not a requirement.

If your boss cares about the customer, then you caring about the customer helps, but often it doesn’t matter much as everyone makes it out to be. On top of that there are way more direct avenues that have a bigger impact on your job development than whatever a user thinks.

[dustingetz]

companies do not really have interests, individual people have interests. The founders, the board, investors, executives, managers, ICs, consultants, vendors, customers. Everyone is doing what their manager (customer) tells them to do while capturing enough value for themselves to make it worth it, recursively. This becomes especially clear when interests diverge, the executives screw the employees, the board screws the founders, the founders screw everyone, the managers accumulate unnecessary headcount so they are responsible for more cashflow, the ICs inflate the estimates, the PMs prioritize visible short term features only.

I also recommend Gervais principle as mandatory read for young managers who don’t yet see the game for what it is.

[mentalgear]

So you're defending the standard company evaluation metrics in theory, yet in your closing paragraph you state from your own personal experience that the outcome aligns with the above commentor's statement (which you disagree with).

[yodsanklai]

I'm saying it's possible to evaluate employees based on some metrics and objective criteria, and it can be much better defined than "make manager happy" in some arbitrary way. That's my disagreement with above commentator.

My experience is that it can be hard to define criteria that align with company's goal (make shareholder happy).