I've seen a lot of posts recently about how to go about hiring hackers. Well not just recently, it seems to be common thing people ask, because it's one of the hardest things to do. You're using an interview, or even a series of interviews to try your best to determine if the person is going to be up to the task.
The problem is of course, you're really just determining if the person interviews well, or if they're the graduates that they've got good marks. The real problem is of course neither of the two matter because neither of them have much if any bearing on how good a coder a candidate is going to be.
The problem is exacerbated in a startup because unlike some of the more successful bigger players you really can't afford to take people on a trial basis for 6 months. What can a startup do?
There are a lot of universities that run placement programs for their students, programs allowing them to get real industry work experience. I've also worked in companies that made it a policy to hire students for internships then offer the best jobs once they've graduated. It was a great way to get talent.
What if you need to hire to fill a more senior role, or you're not hiring at the right time of year? Student internships usually start once at the same time every year so unless you're looking for fill a role then, you're shit out of luck.
I've toyed with the idea of a 'hacker camp', getting all the candidates together, and paying them to work as consultants on a project - then hiring the best into permanent positions. All the candidates need to be free to work on the project, and a lot of people would be pretty scared to leave a job to take part and some jobs would prevent them from working on the project, so running it over a weekend might be a good idea, but it limits the scope of your project. You'd also still need to have interviews, to make sure you were getting candidates who were really interested in the job, and also to weed out any candidates you knew were totally unsuitable.
I think it's important that a 'hacker camp' should produce something at the end of the weekend and you should get as much of your existing staff involved as possible. To make the cost easier to swallow you could even go so far as making the deliverable a new feature or part of your existing product but that's not as important as actually delivering something, because that will be the bottom line when it comes to integrating them into your team.