The story behind Spotify’s big redesign

I spent yesterday at Spotify’s offices in downtown LA hearing CEO Daniel Ek and other execs pitch the app’s biggest redesign in years. The headline is that it introduces a TikTok-like visual feed to the main tab for discovering new music via clips, podcasts, and, yes, video. “This is about bringing Spotify to life,” Ek said during the unveiling.

Given the attention on Spotify’s changes this week, I’m trying something new and devoting most of today’s edition to an interview I did yesterday with Gustav Söderström, Spotify’s co-president who oversees all of product and engineering. He surprised me with the reason for introducing a visual feed and what it’s optimized for. He also explained the reason for changes to the company’s structure in January that saw him elevated to the role of co-president alongside Alex Norström, who manages the business side of things.

My full conversation with Söderström will air next week on Decoder, my boss Nilay Patel’s excellent podcast. In the meantime, I’m giving you the exclusive first look at the highlights of what we discussed. (My colleague Ariel Shapiro also teased part of the conversation earlier today in Hot Pod, her must-read newsletter about the podcast industry.)

I also have some reader feedback from last week’s edition, SXSW tips, people moves, and more. But first, onto the big interview…

Significant redesigns of popular apps are always controversial, and the one Spotify announced this week is no exception. My mentions are filled with people who hate the idea of Spotify chasing after the TikTok UI craze that’s sweeping across the rest of consumer tech.

According to Spotify’s Gustav Söderström, introducing a feed isn’t about going all in on video or even trying to juice time spent. Rather, it’s about embracing a new interface that can make Spotify’s recommendations, which are primarily geared for music, better over time. “I think companies that don’t have an efficient user interface for a machine learning world are not going to be able to leverage machine learning,” he told me yesterday.

Competitively, the shift toward a visual feed puts Spotify in more of a direct position to go after the music discovery that’s happening on TikTok, though I’m skeptical that will work given that Spotify isn’t a social network. Regardless of what happens, the introduction of the feed is Spotify’s biggest move yet at trying to become a true discovery platform for not just podcasters but also music artists, 25 percent of whom self-publish to the service already.

As someone who has been with the company for 14 years and oversaw the launch of its first mobile app in 2009, Söderström has helped lead Spotify through several big shifts. After talking to him, it’s clear that he sees AI as the next big one and that what Spotify announced this week is rooted in that belief.

Here are parts of our interview that have been lightly edited for clarity:

Gustav: I think about it this way: when we ask users what they love about Spotify, they say things like personalization and so forth. But if you drill down, what they really say is that the more new things you show me, the more new things I find, the more I’m going to like this product. So it seems like the love for Spotify is very correlated [and] hopefully causated to the amount of discoveries we drive. You ascribe value and love to the thing that helps you discover new things and have a better life. So discovery is kind of the lifeblood of Spotify.

If you take that lens, then you should expect us to try to do everything to improve discovery. And the truth is, what has happened the last few years with these feeds that autoplay content… It has sort of converged on what is the most effective way to evaluate a new piece of content.

One way to think about it is you need a user interface that can see very effectively what a user likes. And so if you imagine that you have our old homepage where we would show a cover art for a song, you scroll past that. The algorithm can’t know if you even saw it. So maybe you saw it. Maybe you didn’t see it. Maybe you saw it, thought about it, evaluated it, and don’t like it. But it [the algorithm] can’t know, so it’s going to have to show it again. It’s going to have to repeat it a lot.

If you take one of these cards, what happens is when you see this, not only does the algorithm know you saw it — it even knows you heard it and can know that if you listen to it for a while and continued, you should never see it again. So one way to think about it is that it’s like giving the algorithm glasses. The secret of why some of these products are so good at recommendations is not actually that they have better algorithms. It’s the same algorithms with a more efficient user interface.

Gustav: We’re doing something very different. We’re not optimizing for time spent in the feed. We’re actually optimizing for how much you listen to later or how much you save for later. This may sound a little bit like a cop-out, but it’s actually true. The way to think about it is to look at the incentives.

First of all, we’re mostly a subscription business. So we make money from retention, and you get retention when people listen to these long things. Even in the free tier, we actually don’t make money in the feed. We make money when someone later listens to that song and there is an ad in between that and the next song. So our incentives are not to keep you in the feed. Our incentives are to let you evaluate lots of content and put it in your library so that when you get to that background moment, you have tons of things to listen to. So that’s what we want to achieve. We would like you to quite quickly get through this feed because we don’t have a lot of foreground time in Spotify. We’re a background application.

Alex: I’m curious if the TikTok discovery phenomenon is a competitive threat to you all. I would have thought that it would be symbiotic because you find an artist through TikTok but then you go to listen to them on Spotify. You’re trying to have both of those experiences now in one app. Is that fair?

Gustav: You’re using the exact right word. The way we thought about it internally, TikTok, YouTube, they’re actually symbiotic. Most of the foreground discovery has happened there, and fortunately for us, so far, we still get almost all the background. Listening right from that discovery trickles down to us. So it has been and it is symbiotic. But the answer is yes. We feel that most of the time, you should also be able to discover on Spotify.

You need to invest a lot more in machine learning and safety. There’s a lot of cost coming with being one of these platforms. And we decided to take that leap about a year and a half ago. And what you’re seeing now is really us saying here it is, the platform where new artists can break. You don’t have to already be big to get an audience here.

Gustav: Part of the reason for the [Spotify] layoffs and the org change is the macro climate in technology, where funding and money was incredibly cheap for a very long time. People might complain that technology companies were exuberant. But the other way to see it is that when money was cheap, it made all the sense in the world to be very aggressive and grab as much opportunity as possible. And so did we. We got to half a billion users and number one in music and podcasts, and so forth.

But the world changed. That’s not true anymore. Money and capital was very expensive. So Spotify has to change like everyone else. We need to become a much more efficient company. We need to focus in addition to just being a great product. I mean, we’ve always known that the only way to build a great product for the very long term is to also be a great business. You can loan money for a while and build a great product, but long term, you want to be self-sufficient. There is no job guarantee except being profitable.

Gustav: I talk about two kinds of org extremes: Amazon on one side and Apple on the other. This isn’t necessarily true, but if you stereotype a little bit, Amazon is known for parallel teams. You run in parallel. You divide and conquer… You’re not blocked from reaching the consumer. That also results in a consumer experience where you might see three search boxes from three different teams on the screen at the same time. But it works. It’s a trillion-dollar company. There’s nothing wrong with it, but it doesn’t optimize for a simple user experience. It optimizes for speed.

Apple is the opposite. They ship much, much slower than Amazon, but no one gets to put their own search box there. It is centrally synchronized. So they manage to build something very complex that still feels like it was built by very few people for a single user. So we chose to adopt more of that. We needed to synchronize the company and sort of eat that complexity of music and podcasts and audiobooks and potentially other things for the user instead of just shipping our org chart to the user.

So we built this org where we have three horizontal layers. We have a platform layer [called] Spotify technology platform. We have an experience layer called the Spotify experience layer — which is all the applications, all the surfaces, mobile apps, cars, desktop, and so forth owned by a single person — and then a personalization layer which owns all the personalization.

It’s a synchronization function that actually slows things down [internally]. But then these people are forced to eat all that complexity that would otherwise end up with the end user. And so these sort of vertical businesses — the podcast business, the music business, the audiobook business — they can’t actually just go and ship stuff to the user. They have to go through these synchronization functions.

That was the big change that we haven’t spoken much about externally, and it is all in service of being able to do this without drastically increasing the complexity for users to keep it simple. So far, we think it’s working. We have quantitative metrics that say that it’s working, that consumption on Spotify is way higher than any competitor.

You can’t both let everyone run fast in parallel and be simple at the same time. I often say that people come and ask me for, you know, more autonomy, more swim lanes. They want to have more responsibility. And this is a very natural thing in companies, especially American companies, I find. And what I say is, “No, we’re not doing competitive swimming. We’re doing synchronized swimming.” A much harder sport. Much harder to execute but much more beautiful to look at when it works.

An OpenAI employee responded to my tip that AWS isn’t able to meet demand from startups wanting to launch AI features: “The thread to pull on is confirmation of Nvidia’s annual GPU production count and allocation across the cloud oligopoly.”

“Horizon might be the only app in the world that has more reporters covering it than consumers playing it.” - A VR startup CEO

Speaking of Austin, I’m not headed to SXSW this coming weekend, but a bunch of my colleagues from The Verge and Vox Media are. Our Verge team on the ground is doing a live Vergecast this Sunday at Slack’s house, and Kara Swisher will be interviewing Kevin Systrom later in the day in Ballroom D. Another talk I’d try to make if I was there is Saturday’s interview with OpenAI president Greg Brockman.

If you like good Mexican food and drinks, I highly recommend reserving a table at Suerte. And if you get sick of BBQ, Eater has a great guide to the best sushi in Austin.

That’s all for this week! I hope you enjoyed this edition. Please don’t hesitate to forward it to a friend.

I’ll be back next Thursday. In the meantime, if you have any feedback, tips, or suggestions for how I can make this newsletter better, let me know.