Leaving Spotify
All good things come to an end, and as of today, so has my time at Spotify.
Despite occasional complaints about various aspects of the work, overall, I enjoyed working here. If I’m being honest, advertising still wouldn’t be my first pick for a product domain, but I’ve been pleasantly surprised by the sophistication of its technology and the interesting work that can be done. I also appreciated how my work challenged me to reach beyond the familiar world of web UI and learn new things. And, of course, I was always grateful to work with so many amazing colleagues.
Even though I thought about switching jobs from time to time (yes, I did a few interviews and even passed some), I didn’t really expect the end to come so quickly—it’s been barely three years since I started here! So it is with a heavy heart that I part ways.
So long, Spotify, and all you wonderful people at Spotify! I wish you all the best!
* * *
Every chapter’s end is an opportunity for reflection. I will keep it short and sweet this time since I’ve already shared my thoughts on my first and second Spotiversaries. I don’t have any special wisdom to offer anyway. Everything I’m sharing here is already known to anyone who has worked at a large engineering org. Still, these lessons are noteworthy to me now that I’ve lived through them myself.
So here they are, some lessons learned from working at Spotify Advertising where we had a handful of large, ever-changing codebases with contributions from multiple teams:
-
First impressions are very, very important. This is especially true in the context of a large org where we don’t interact with each other so closely so often. We are all rightfully occupied with our own tasks and priorities. So we rely on shortcuts—like our memories of earlier interactions or stories we hear from other coworkers—to size people up. And because we’re all busy, we rarely think twice about who this or that person is—the rare exceptions to this are largely reserved for immediate teammates we interact with on a day-to-day basis.
-
Alignment matters more than getting individual tasks done. Alignment comes in different shapes: technical decisions, product requirements, timelines, assigning ownership and responsibilities, and many more. This is not to discount the importance of writing good, well-organized, and well-tested code every day. But from the organizational standpoint, at a certain scale, the relative significance of good code gradually declines. A full-fledged feature launch that took several months—even years—of R&D time could be rolled back in the snap of a finger due to misalignment. All the carefully crafted code for this launch then turns into tech debt to be cleaned up.
-
Even maintaining mediocrity takes effort; fighting it requires conviction and coordination. In a large org with ambitious goals, it’s almost inevitable that many of us work on multiple disparate projects at any given time, where most projects require cross-team, cross-functional collaboration. When combined with aggressive timelines, these conditions often push us into a constant MVP state: we’re incentivized to do just enough to get one project across the finish line so we can move on to the next. Simply keeping our product from deteriorating demands discipline and determination from each and every IC. Making any significant improvement takes leadership and prioritization on an organizational level—this fight against mediocrity comes at a nontrivial cost, too.
* * *
What’s next? I’m lucky to be moving on to a new, exciting adventure. Yes, it’s advertising again, but it’s a rare opportunity to work in a young, fast-growing org at a major global brand. My work will focus on a different stage in the ads pipeline, which means plenty of opportunities to learn something new. What makes it even more exciting is that I’ll be joining several familiar faces from Spotify, including a couple of old teammates!