[nicolashahn]

I'm a SWE at Meta and do technical interviews. I'd say you're right on most things except down-leveling. There are some very specific things interviewers are looking for in the behavioral and system design interviews and I believe they accurately predict how an engineer will perform at a given level within Meta, which may be different than how they would at the "same" level at another company. In my experience and from speaking to other FAANG engineers, Meta expects engineers to do a lot more than just code past E4. E5+ ("senior") should be able to do some of what a PM does, generate impactful projects on their own, and be able to be a team lead (although that's usually for E6+), plus some more stuff. I know at least I was definitely leveled fairly, and even struggled a bit to get used to the expectations when I first started out.

[leeny]

Thank you for responding! Would you say the leveling criteria has changed in the last year? I ask because, while the current criteria could very well be legitimate, candidates weren't down-leveled nearly as often until the downturn.

[nicolashahn]

There's no official policy or guidance that has changed on the expectations for the behavioral/sys design interviews I know of, which is fairly transparent. I saw your stat for 55% self reported down-levels, do you have a stat from before the downturn?

One thing I have noticed is that in my experience (around 70 interviews total since I started giving interviews in 2021) the average performance on the coding portions dipped sharply around the time that our stock did. My theory is that the more talented candidates focused on companies with a better outlook and so it was weaker candidates that were making it to the technical interview stage. I've noticed that in the last few months I'm seeing candidates do better on the coding interviews which supports my theory.

[leeny]

I wish we did. Honestly it's not something we collected because we didn't hear about it happening from our users.

That's a cool hypothesis. I think we can actually pull something similar: grab average performance in Meta-themed mock interviews on interviewing.io over time and graph it against the stock price.

[leeny]

Here it is: https://docs.google.com/spreadsheets/d/e/2PACX-1vSQBAZOChbsq...

There's something there... but it's not a very strong connection. Of course, mock interview performance isn't necessarily in lock step with how all candidates are performing in real interviews at Meta.

[thedevilslawyer]

Thanks for bringing data. You're being a mensch.

[mandeepj]

> since I started giving interviews in 2021

You meant taking interviews?

[patothon]

maybe it's the other way around? other companies started to up level during covid and hiring mania?