I have had a very recent internship interview (actually about four or five) with Joel Spolsky’s Fog Creek software. Unfortunately I must say that I am not impressed with their at-best repetitive interview process.
Let me describe the painstaking interview process I went through. It started really awesome, then things gradually declined – you will see what I mean. Our school has already a default application process for applying to internships. Of course, Fog Creek will not have any other way except their own. First, I needed to send an email explaining why I want to work at Fog Creek, as well as attaching my resume. This is, according to them, the first of three steps of their interview process. Immediately after sending the email I in fact received a funny automated response. The second step is a phone interview (of course, they insist on going outside of our standard phone interview procedure for our school and want to set up in their own way). I interviewed with one of their engineers. It was just an ordinary interview. I described some of my past experiences and we proceeded to a technical question. Its about how to design a data structure. We finished with with a QA session.
Apparently I did well enough so I get to participate in the last phase. Fog Creek flew two engineers (one of which I interviewed for) and a PR person all the way from New York. Apparently they think they are the most important people there that day and decide to schedule 3 hours of my day for them without any room for freedom. As a direct consequence I had to miss some of my classes just to attend these interviews. The first was a group interview. We spent one hour sitting there listening to them cracking jokes (I think they thought they owned the place) and talking about New York. At this point I am just plain annoyed. Next came two hour-long interviews with their engineers. Guess what, two more technical questions. They asked one already. Now they are asking two more, which in my opinion adds no value since they are similar in nature (they do actually get to see you write code – but they didn’t need to do it twice). I ended up spending time writing C code on how to dynamically reallocate an static array when its full (which is a small part, but not all to the question – I am hesitant to reveal what the actual question is here). I think I did well enough (hey, if I were to name one thing I am really good at, it is to answer these questions during interviews). I think I managed to finish one of my interviews at the 30 minute mark while the person before me was going overtime.
They managed to select two out of the three candidates, including me for a third interview. This time Joel flew in, a few days later, to do the interview himself. I must say he is quiet a character. Definitely someone capable of articulating his thoughts clearly. I had being reading his Joel on Software blog the day before out of pure boredom, and I admit I found it quiet engaging. So we chatted for a bit, talked about my previous experiences, and then proceeded to move on to yet another technical question (instead of a design or a systems question). It is a little bit different, in that he used “his own language” which is essentially Scheme. I had about four weeks of Scheme experience in a course three years ago so I did know the basics. I didn’t have too many trouble, and finished again around the half-hour mark and was expecting another question. But he didn’t have any (I guess he deviated from the list given above). He then told me it was now time for me to interview him and ask him any questions I may have. Its all good, except I didn’t have any questions! Remember how I had three QA sessions with his engineers (and they told me about Wasabi, and how one of the interview questions stemmed from on how some Office component was implemented when Joel was working there – isn’t this protected secret or something?) and I had that stupid one hour long group interview talking about nothing (but his company and New York). So I told him that I found his blog to be interesting (by mentioning one of its recent entries) and told him I had no more questions. I was hesitant to throw Jeff’s commentaries about Wasabi to him because I was sure he would not like me to question his authority. I have a feeling he didn’t like this very much. You know, it was a neat line “now its time for you to interview me”, and what I did was essentially shooting it down completely by not asking him any questions about Fog Creek. he even leaded me in by asking “do you have any questions regarding the job itself?”. I told him discreetly that I learned a lot about his company during the group interview and did not have any more questions. In the end, the result was that I got rejected at the end of the day as seen from our school system.
But that’s not all. Also at the end of the day, I received an email from one of the advisers. Apparently someone gave him feedback that I did not appear enthusiastic in one of the interviews and apparently I did not do research on the company. While I can’t be 100% sure it was Joel, I had compelling reasons to believe that was indeed the case. I immediately dejected this feedback and portrayed myself as “offended” at such a “blatantly false accusation” to the adviser. I also made it clear [to the adivser] that I was not about to be enthusiastic about some enterprise content management or bug track software that I never used or will ever use in my life. While Joel’s remote assistance idea seemed to be something I could play around with, the fact that I had to pay money to use it beyond the two minute mark completely discouraged me from trying the the first place (this all counts towards research btw). There is one misconception about every job posting. The interviewer always expects the candidate to be enthusiastic about the position. Thats not going to happen with me. never. You want me to be enthusiastic? Give me an offer and give me something interesting to work on. Otherwise we can both move on. I never make false representations of myself during interviews to appear enthusiastic. The only time I will be interested is when I genuinely feels that way – like if a neat idea is rightly so, I’ll do it.
So in the end, after wasting ten hours (not counting the time I’ve spent writing this thing or many emails to that adviser) I did not end up with anything other than a reject from Fog Creek. But then again, I had seven other offers to choose from, and every bit is as bit as good as Fog Creek.