Evaluating if a possible employee-employer match is a good fit often requires several hours of idiosyncratic effort from both parties, both parties usually want to do a bit of due diligence. A common outcome is that the mutual fit will not be a good one.
If a platform makes it very easy for people to apply for jobs, and very easy for jobs to be advertised, but is not good at matching and filtering the number of possible matches to a small number of high-quality matches, then both employers and prospective employees may get flooded with a large number of low-quality matches that have zero or a very low chance of resulting in a deal that is agreeable to both parties. That doesn't really help anyone.
From an employer's perspective, perhaps what they want is identifying some channel or way of addressing and reaching a particular sub-population* of applicants who are more likely than other populations to be a good fit, to make their hiring process more efficient. Would you prefer to get 1000 applications for a role, where each person is 1% likely to be a great fit, or 10 applications for a role, where each person is 30% likely to be a great fit? In the former situation you expect to have 10 great candidates, and in the latter situation you only expect to have 3 great candidates, but the latter situation requires one or two orders of magnitude less money and effort to process the applications. Similarly from the perspective of someone looking for a job. Maybe it's something of a marketing and statistics puzzle for both parties.
It may not be an advantage to either party to participate in a large central marketplace that makes it harder to differentiate themselves from their competitors, as it commodifies both sides to some extent.
* i mean "population" in the statistical sense.
