The offer finally arrived from the company I'd been targeting, and the number beat what I was expecting.
Over a year ago I landed a Spotify interview by applying cold, which surprised me, since my background is a good (though not target) school and an employer people know but don't think of as big tech, and certainly not FAANG.
Apart from internships, I had never been through a proper interview loop before. For prep I did online coding practice, an online resource for system design (I'd never studied that topic before), and YouTube videos for behavioral questions. I was very stressed, yet the feedback I received was strongly positive. Then the role was closed, which hurt even more after I was told I would likely have gotten the offer. A few weeks later I had calls with two other hiring managers, and sadly neither of those led anywhere.
About a year on, a referral got me another interview, and the process was laid out just like the first one.
- Recruiter screen: Introduce yourself, explain what draws you to the role and to Spotify, plus an overall vibe check. I think it works best to be yourself and treat it as a light behavioral conversation.
- Technical screen: It began with simple questions about programming language concepts, networking and backend services, then a short discussion of a project. After that I was given a twist on LeetCode 3 (Longest Substring Without Repeating Characters).
- Final rounds:
- System Design: The task was a service that stores playlist cover images. A person has to be able to upload a picture for a playlist and change it later, and anyone should be able to load those pictures, for instance inside embeds on other sites. It also needed to plug into the services that already existed. The interviewer then asked how an auditor, say, would use the system, and what should happen when an upload or the processing step breaks halfway.
- Coding: A variation of LeetCode 1152, Analyze User Website Visit Pattern. There were quite a lot of follow-ups on how the runtime behaves for different inputs (this problem has several inputs that can vary, not only N), and on what I would change in style and function if the code were going into production.
- Behavioral: This round surprised me most, because I'd assumed it would resemble the questions from my earlier cycle. Several were about AI: how I use it and how I've encouraged my team to adopt it. I hadn't prepared for that, but staying current on AI in my full-time job and side projects bailed me out. The other questions were the usual behavioral kind, just tweaked enough to throw you. Rather than asking about resolving a conflict with a coworker or manager, they asked about a time I made the team more positive. My advice is to be yourself and really take a moment to think before you answer, because the interviewer seems to shape follow-up questions around how the conversation has been going.
- Case Study: Debug a production server while playing the on-call engineer. I'll keep the details vague, but staying curious, asking questions and drawing on your own experience to pick an approach seems like the way to go.
Overall my impression is that Spotify really looks for people who fit its culture. The interviews do have technical parts, but being yourself, communicating clearly, staying relaxed and working collaboratively will get you much further than at companies that lean harder on the technical side and make it tougher.
I ended up with an offer at the top of the pay band, which felt great. Total comp came in just above $230K for a mid-level backend role.
Discussion
Loading comments…