It really isn't that expensive to send a couple of engineers to a school if you're hoping to hire a couple of talented students, and maybe some interns. You really only need one event a year.
Except that there are hundreds of schools that fit this description. Sending recruiters to any one of them isn't that expensive, but sending recruiters to all of them is quite expensive.
Which isn't to say that there aren't students there worth recruiting.
What about putting those schools in a pool and selecting from them randomly, eventually weighting them by historical success if/when you have enough data?
If recruiting is a competitive advantage, why not be innovative and data driven about it? If it's not a competitive advantage, then why spend time and energy on it?
