A company doesn't need to know they are under compensating for a role from a third party. They'll find out by the quality and number of applicants pretty fast. I've seen this in practice directly when we couldn't hire for certain roles, raised the range, filled the spots.

On the other hand, to know that other companies are paying lower and still able to deliver roughly the same work is harder unless you know how much they are paying.

I have to disagree. If you're not getting quality applicants, how do you know if that's because of your salary range, the default applicant pool, or something idiosyncratic to your company?

If you're a new startup founder, you don't always have a good sense of what the default applicant pool should look like. You might have a sense of what quality looks like but how would you know without recruiting experience what the mix of quality to non-quality applicants is supposed to be? There are many reasons why you might not be getting the number of quality applicants you want, and compensation is just one of them. Salary benchmarking data helps eliminate that as a possible cause.

Because when we raised the range it fixed the problem.

I'm not talking about you, I'm talking about people in general.

You have a strange view of the world. People have limited hours to sell for money, so other than undifferentiated work (where you don't get recognized for performance), workers compete for the best jobs, and so more skilled workers will on average make more money. Another factor is that people do lots of shitty things if the money is enough. Salary is the most effective way to attract and retain people, other than if the company has dystopian practices (and even then...).

How do you explain people moving to the USA for better salaries and crap on European salaries when Europe affords workers more vacation time, more social nets, healthcare, etc? Because the USA pays more money.

I think there’s a pretty big gap between “salary is the most effective way to attract and retain people” and “the only possible reason you’re not getting as many quality applicants as you might hope for is salary”. As a very basic example, your standards for quality may simply be unrealistic.

There's a big gap between those two statements but I made neither of them.

Uh, go back and read your post. You certainly made the first one :)

I think it's a pretty safe assumption that a company paying to gain information that gives them an advantage in negotiation isn't going to freely give up that advantage.

You're assuming that the only reason to pay for Pave is to get a negotiation advantage. My point is that there are other reasons, for example, to make sure that you're not below market.

I've never run a company and probably have a chip on my shoulder, but I also think that it's a reasonable assumption that most employers want to pay as little as possible.

I guess my point is that without some sort of sense of the market, whether through Pave or something else, the motivation to pay as little as possible may lead some employers to have lower salary ranges than they would otherwise.

Obviously, this sort of information would always be used for good.

Because information has to be used for good or bad and never both?