I suspect it is a similar ratio for any software company.
My best hire ever was a math teacher in a previous career that wrote his own tools. Worst hire ever was a guy with a PHD that thought he was still in university.
The worst hire I made also had university credentials and apparently wasn't educated in how to deliver anything but academic noodling.
(I even worked there briefly for six weeks. But that's another story.)
(But I was at eg Google before at some other point in my career.)
- the same unsuitable candidates apply many times. This is not so surprising: suitable candidates get jobs and so stop repeatedly applying
- perhaps many people can be hired for a single ‘open position’, eg perhaps one is ‘software engineer based at headquarters’
- if many applicants are from people who are grossly unsuitable, eg maybe you must demonstrate job applications to get some state unemployment assistance
- maybe Meta aren’t hiring much at the moment and so it is harder to get in than other companies that are hiring
- the number might be misremembered
The downside to applying is trivial so even if chances do seem low, it can still be reasonable to apply. I think the probabilities implied by the comment I’m replying to are so extreme that you should have some higher estimated success probability from also considering the chance that the comment above is misleading.
The thing one often sees about automated cv filtering systems is mostly widely-believed nonsense. It is a story that is told to job seekers to sell cv-optimisation services. That’s not to say automated filtering doesn’t exist – a simple case is filtering out candidates that were recently rejected – but I think even if a lot of people are filtered out, that doesn’t mean you will be.