One of the reasons I wrote the blog post I linked in my previous comment is to tell a story, in concrete terms, of why this kind of "well why don't you just" story fails in all sorts of ways both obvious and not. It's thousands of words long, and it doesn't even cover all the big points of how complicated actually trying to build something better is. It omits many things I spent weeks thinking about in the moment, and barely mentions people who worked on the problem for years, in the interest of being only a twenty-minute read, in an effort to just give a taste of the complexity of working on a problem in its full complexity.

----

> job listings/postings with a bit more integrity behind the them. Submit something, get an acknowledgment

What kind of "acknowledgement" are you looking for?

The vast majority of applications, anywhere, suck. Some platforms suck more than others (and LinkedIn is one of the worst), but if you've never done hiring yourself, I'd like to impress on you just how bad the average applicant is. Even on places like Work At A Startup or Wellfound or direct apps on your company website (which tend to be higher-quality on average), most applicants are awful. And thus, the acknowledgement will almost always be "nope sorry bye now".

If that's the acknowledgement you're looking for, lots of places do do that (in our case, we'd close out an app if not actioned on in a short period of time and send a message saying the company hadn't acted). But are you really going to use an entirely different platform because they say "no" rather than letting you infer it, even if that platform has 1/10th the jobs or takes a bunch of energy to sign up?

----

> perhaps limit number of applications to reasonable numbers

We did this. We did the same for recruiters, actually (limiting the number of messages they could send candidates).

In practice, this didn't make the average message significantly better (granted, this was pre-GPT), and it created a new set of problems. Recruiters would get pissed at us if a candidate didn't work out, because they hadn't just spent a minute sending a message, they'd spent one of their precious message credits. They stopped being willing to consider borderline candidates or candidates with non-traditional backgrounds for fear of wasting messages (even though in the vast majority of cases they weren't exhausting their limit! FOMO is a thing, and user behavior is frequently irrational or uninformed!).

Nor did candidates put meaningfully more effort into the apps they were sending out. Nor did recruiters meaningfully value applications any more than they did elsewhere. Unless you're doing very vigorous tracking, the difference between applications being 99% crap and 97% crap is not very noticeable and requires rather large sample size to detect. That's a 200% improvement that can go totally unnoticed (and good luck achieving that much of an improvement in the first place).

That's even assuming your users are statistically-rational, which of course they are not. Recruiting is low-information, high-emotional-valence, low-trust, low-sample-size, low-agency, and high-noise, and that's about the worst possible mix you can stir up for cognitive biases of all kinds. Most recruiting advice is one step removed from "you should message candidates with a Q in their name only on tuesdays under the first full moon of the harvest" as a result.

And that's even setting aside the adverse-selection effects that mean alternative platforms tend to end up dominated by (a) hard to fill roles and (b) desperate candidates (since those are the ones that don't get matched through default channels).

And yes, you can come up with solutions for these things - when you know they're problems. But those solutions create problems of their own, often in unexpected ways.

----

> It's an overloaded marketplace optimized for engagement rather than good outcomes for either party.

This is true, but what's implicit in your argument here is the claim that "being a marketplace optimized for good outcomes" is in fact the primary criterion for being a successful platform.

I'm not saying LinkedIn is good for job searching. It isn't. I'm saying that "the existing solution is bad" is not sufficient to draw people to alternatives by itself, and that building a better solution is harder than it seems.

I don't mean to say it can't be done, or even that people shouldn't try. I'm trying again (just not with a platform), so obviously I think there is some degree of solution to be had here. I just think that this kind of comment betrays a kind of "one simple trick" attitude that just is not engaging with the complexity of the problem or with the fact that incentive structures often do not point to "make a thing that doesn't suck".

That said, I did learn some time ago that you can use various search operators (Boolean, parentheses, and quotes) which make it a little easier to actually get relevant job search results.

In general though, applications are always a terrible way to find a job. They're a last resort when your networks and other connections and "ins" fail.