A friend of mine had recently started working at Airbnb and they put me in touch directly with the hiring manager of the data science team. A referral is the quickest way to get an interview; I had applied to Airbnb weeks earlier and still hadn’t heard a peep.
I had a pleasant half hour chat onsite with the hiring manager, who offered the option of coming for an onsite tour or a phone interview. It went well enough, and by the time I’d left the office, we’d tentatively planned a full onsite interview for the following week.
The on-site interview was to be structured as follows: in the morning, I’d work on some basic data manipulation tasks under the guidance of the 2nd-in-command of the data science team. Following that, there’d be a half hour chat with him and another senior member of the team. Then, lunch with the whole team. In the afternoon, I’d have 3 hours to work on the “challenge problem”, an open-ended question in which I’d go through their past data to try and answer a question related to running their business more smoothly. At the end of the day, I would present my findings to the whole data science team.
When I showed up that morning, Mr. 2nd-in-Command wasn’t available. The hiring manager himself, who seemed annoyed at having to intervene, got me set up at a standing workstation just a few feet away from where the data science team sat. You have the choice of using STATA, R or MySQL. After an hour or so, I’d finished the initial problems, and it was time for the chat with Mr. Second-in-Command and the other senior team member. Both were busy with meetings, so the hiring manager suggested I get an early start on the “challenge problem” before lunch.
Working at a standing workstation got exhausting after a while. I asked the hiring manager for a chair so I could rest, but he never mentioned that by pulling a lever, the standing workstation could transform into a desk I could sit at. I didn’t figure that part out till mid-afternoon, when I saw someone nearby did the same.
Airbnb’s data science team consisted of about 10 people. I found them to be extremely bright and a bit introverted. On a couple occasions during the afternoon, I walked over to the data science desks to chat and see what they were working on, but no one seemed to take the slightest notice that I was there. I asked Mr. Second-in-Command about the “chat” we were supposed to have earlier in the day. Looking as if I had blindsided him, he checked his calendar, then walked over to my standing workstation, asked me a couple questions about my background and whether I had any issues with the data challenge, then proceeded back to his desk.
The “challenge problem” was simple in principle yet challenging, since there were so many ways to approach it. You only have 3 hours to do it, and it can take a good chunk of that to read in and understand the data. I wasn’t happy with the approach I took, but as the clock ticked I felt too committed to change anything. In hindsight, it would have been better to take a step back and think more creatively before diving in and trying the first thing that came to mind.
At 4:30pm, Mr. 2nd-in-Command rounded up the data science team for my presentation. 8 people attended, not including the hiring manager, who was busy in meetings. 4 people brought their laptops and buried their noses in them and I had to fight for their attention. The team didn’t seem very impressed by my presentation, and by the time I was done it was clear I wasn’t getting any further in the interview process.
I was disappointed in the day. Part of that was feeling like I could have done better on the data challenge, but I also felt let down by the process. It was less of an interview, and more of a skills test under pressure. In a way, that’s a good thing because they can see what you’re capable of. But it also feels like they’re giving you the cold shoulder as a candidate, as you’re spending all day on the skills test, and don’t get much of a chance to talk to anyone on the team. Lunch is only 30 minutes, and your mind is busy trying to think of solutions to the challenge problem. It also doesn’t help when neither of the 2 people who are supposed to interview you bother to do so. Given how little time I spent talking to people, I wonder why they don’t just give the skills test remotely, so you don’t have to take a day off work to do it.
I know it’s a data scientist job and it’s mostly technically focused, but I was hoping that Airbnb would have been more fun to interview at. When I asked the team what they did as a team for fun, their reply was that generally, every other team besides data science was more fun to hang out with, and they hadn’t done anything as a team in a long time.
It took 2 weeks for a recruiter to get back to me. Although an offer from Airbnb would have been hard to turn down because of their success and reputation, part of me wonders if I would have enjoyed it.
