On Technical Interviews And Diversity

# COMMENTS

I recently stumbled upon Laurie Barth on Twitter and noticed a couple of tweets about tech interviews and the hiring process, something I've been thinking about recently.

I guess I've been thinking about it because I'm teaching a section of Data Structures which, along with Algorithms, provides the base knowledge you need to have a shot at the typical "Google Style" technical interview.

I say "base knowledge" because just completing those two classes doesn't really prepare you. It gives you the building blocks but you really have to prep specifically for the types of questions they might give you. Unfortunately, for most kids it involves a whole lot of rejection and failure before you start getting offers.

You might get a problem that has a bad brute force solution and during the relatively short interview, devoid of your typical resources and comforts you have to turn it into a much more efficient one. The argument is that the interviewer is really trying to see how you think about a problem but having spoken to many, many tech employees who do these interviews that's a specious argument at best. It's really a test of "who has seen the problem or something similar before but can fake that they haven't the best."

While it's true that you will get some candidates who can solve the problems on the spot you're going to miss out on a whole bunch of great candidates who just don't do well on that "Google style" of interview. At the same time we keep hearing that, as Laurie tweeted, the skills required on the job don't match the ones needed to pass the interview.

This doesn't really matter to the big, hot companies. The Googles and Facebooks of the world. So many students coming out of college want to work for them that they can afford to miss on tons of great candidates so long as they hit on enough good ones.

I put the word "diversity" in the title of this post and you might wonder what I was going for there. So far, I'm just ranting about the interview process. This becomes a diversity and privilige issue because the process favors "elite" CS schools with established CS cultures. I'm not talking about the fact that they pre-select high performing candidates nor the fact that all the tech companies go to those schools, sometimes exclusively for recruiting. I'm talking about how the school's student tech culture helps prepare the students for the technical interview.

From my discussions with many graduates from many of these institutions, while classes don't prep for interviews, students at these CS schools regularly share interview questions. Friends and classmates talk about the questions they got and help each other out with practice sessions - formal or otherwise. I'm certainly not faulting students for doing this but it's another known secret. We all know that the more tech interview questions you see the better prepared you'll be - both because you'll get better at that specific problem solving skill and also because you'll likely get interview questions that are already in your library. It's that latter part that makes the whole process very nod, nod, wink wink.

Nothing stops other students at other schools from doing similar but it's much harder at a commuter school where kids live off campus and have to work part time jobs. It's the reason I'm fighting so hard for space and working on building a postitive tech culture at Hunter. Not specifically for interviews but to help my students become a support network for each other.

I'll finish by noting that while this "technical interview" process might more or less work for the big boys - the hot company du jour, it probably doesn't for eveyrone else. A buddy of mine was talking to me about the process he uses at his much smaller company. He noted that he could use the "Google interview" process for his future employees but since his company was much smaller and lesser known, people who did well would probably also get offers from the big boys and go work for them. His goal was to come up with a process that catches the great candidates that don't do well on the Google interview to try to catch the great talent that doesn't do well on today's technical interviews. A number of smaller companies seem to be starting to do this and from the few I've spoken to it seems to be going well.

I get that big companies are machines and look for uniformity in all processes including the interview process but there has to be a better, more human way to evaluate talent than a one size fits all process where the norm is failure - particularly when it seldom relates to actual skills needed on the job.