You're forgetting something: click through rate. Unless it's somewhere above 90% (in which case you're one hell of a marketer -- or you're paying for the traffic), then you're essentially burning your money.
This is how I found my first job (straight out of college, industry had crashed in south africa after the .99 bomb, I had been looking for a while but no luck). This is what I eventually came up with, you decide if it's how hackers find a job.
- I put an ad in a popular job seeker's rag. It cost a pretty penny, but the end result was worth it.
- The ad was not as a job seeker, but as an employer
- The employer, of course, was fake but I described the kind of company that I wanted to work at
- I then listed credentials and traits similar to mine, for this was the person my false company was looking for (in the end looking for someone like me upped my chances of success)
- I put an email and a fax number, where they could send resume's. I insisted on references being passed through.
- Naturally, my fax ran out of paper, me email box was flooded.
- I screened the resume's of the candidates, paying particular attention to their previous employers, eventually picking the one I wanted to work for
- I then phoned the few candidates selected and asked them about their previous job (all the while still pretending to be a potential employer -- it sucks, but I was running out of hope and really wanted a good job!)
- I finally picked the one that loved his previous job but decided to leave because he had done everything to be done there
- I phoned the reference provided in his resume`
"Hi, this is Wesley. I met so-and-so at a party on the weekend, he mentioned you may be looking for devs that are x, y and z"
I had a full loop with Google last year. It was awful. The guys were really aggressive and incredibly offensive, reminding me time and time again that they hated the company I worked for. Which is fine if it's your personal opinion, but don't invite me to your offices only to insult my employer. Kinda silly, really.
I can't help but wonder that maybe the interview process is broken. Can you really gauge how effective a dev is based on his might at solving a few programming problems on a whiteboard?
I've seen the strongest candidates in this field end up being the worst devs (bad team players, no design docs, not adhering to spec, crap unmaintanable code), and vice versa. Which makes me wonder. Anyone using any interesting techniques out there other than the standard "here's a tricky problem, now solve it"?