The "solution," such as it is, is that companies strongly bias towards referrals and managers towards people they personally know. And, from some conversations I've had, that is exactly what is happening. With the result that it's tough for junior people with no real networks (OK maybe their school is a signal) because companies really don't want to sift all the junk they're getting and I don't really blame them.

>OK maybe their school is a signal

"School" is quite variable.

Weak signal: you only went to class and did OK in them.

Strong signal: you had an internship, or undergrad research experience, or part-time employment as a TA/tutor, or have a completed project to show off, or some kind of non-trivial community/group/club/fraternity leadership.

Really strong signal: you published a paper with someone I know and they recommend you to me.

Typically, undergrads aren't publishing papers. I did co-author a paper in grad school though it was irrelevant to my eventual job.

Most people are looking at whether you just went to (whatever they consider) a top tier university.

Absolutely, it's very rare for an undergrad to be on a paper. But that's what makes it such a strong signal: it shows they had the grit and maturity to contribute to a research effort to completion, in a team with people more experienced than they are. In an interview, it gives them something non-trivial to talk about and be proud of. That's very likely a strong junior candidate.

There are a ton of things you might look at for a newly graduated undergrad beyond grades: research and other academic projects, sports teams, editor on a newspaper, etc.

Why are sports teams important signals?

That's a solution, but I'd prefer a system that, when ideal/efficient, is optimal.

I'm not sure what other solutions look like: Gatekeeping of various forms including institutions and certifications, letters of introduction like essentially the US service academies, standardized tests, informal networks, etc.

There's two distinct reasons why more qualified candidates might get skipped over:

1) There is too much noise occluding their signal.

2) There is a form of gatekeeping going on.

Gatekeeping only really works in exploitative systems (e.g. "me and my children are the masters, and you and your children are the slaves") or when the noise is so high that companies wouldn't gain much from not gatekeeping (e.g. Harvard admissions in the late 1800s).

So, if you don't exist in an exploitative system, providing more signal is going to both benefit deserving candidates and punish gatekeeping companies. I don't see why a reputation score would increase gatekeeping.

At the end of the day, every applicant could be ranked on their ability at the job. Wouldn't it be best for everyone—companies and prospective employees—to know where they rank up, so they don't waste time applying to hundreds of jobs or sifting through hundreds of applications?

The only people who are hurt are the hustlers: people who spend far more time hustling for a position than gaining the skills needed to do well in that position. Their goal is the extreme limit of noise, where success rate is directly proportional to how many applications are filled out, and I have no sympathy for the destruction of the commons (that I have to live in).

For a lot of things, hustling is probably at least as important to me as a hiring manager as rather amorphous "skills needed to do well in that position" at least as an entry-level employee. Of course, I don't want someone who has none of the skills needed for the job in most cases but they probably don't know most of what they need to learn anyway.

The weird thing to me is that I don't see this happening at the large FAANG companies - referrals don't seem to move the needle whatsoever anymore, and not just for me but for quite a few of the people in my network.

On the flipside I'm not finding good resources to find startups to apply to that don't have hundreds of applicants already. There's no good answer the market has come up with as far as I can tell, so everything just gets worse for everyone as a result.

About uni apps in the US: Most people limit the number because there is a modest fee associated with each. I doubt it is 2-3x since last generation.