As some Spotify users find themselves in an Espresso echo chamber, how are streamers discovering new music?
“I know I have good judgment, I know I have good taste”, sings Sabrina Carpenter on her latest single, ‘Please Please Please’. But, after the song’s release, some Spotify users found that no matter their “judgment” or “taste”, they couldn’t play music without the app queuing one of Carpenter’s recent hits immediately after. The supposed phenomenon has prompted Twitter conspiracies and informal investigations but, whether you believe the claims or view them as diverting credit from a young singer’s success, they draw attention to how Spotify is not only used for streaming music, but is expected to know and reflect its subscribers taste through personalized recommendations. With more than 615 million users, who on average spend over two hours a day on the app, Spotify has great sway over which music many people are exposed to. Although streaming makes accessing a variety of music easier than ever, can a reliance on it for recommendations limit our listening?
Glenn McDonald, ex-Data Alchemist at Spotify and author of You Have Not Yet Heard Your Favourite Song, explained that which songs Spotify plays you is mostly determined through “some version of collaborative filtering: find other listeners who listen to some of the same things as you, and see what else those people collectively listen to.” What might further explain the ubiquity of certain artists on Spotify is how these algorithms can be “tuned to give you more or less familiar songs, newer or older songs, more or less artist variety”. In this way, Spotify can help users discover lesser-known musicians, or push chart-topping singles. McDonald has also noticed a trend in “turning more ‘recommendation’ processes over to AI/machine learning systems that attempt to optimize for results (like longer listening time or more songs played) instead of paying direct attention to the music itself.” This favors less intense, ‘background’ music, or shorter songs, rather than aiming to play songs the listener would most enjoy.
“If you don’t want algorithms to feed you passive listening, get active.”
Glenn McDonald
There isn’t a button you can click to avoid these algorithms, but that doesn’t mean they are controlling what new music you discover: “If you don’t want algorithms to feed you passive listening, get active” McDonald says. Within Spotify, he recommends the ‘Fans Also Like’ and ‘Discovered On’ features. Although young people are less likely to discover new music through the radio (Gen Z and Young Millenials listen to the radio 30% less than older generations), they are finding ways outside of streaming to diversify their listening habits.
Though discovering music on social media risks entrapping oneself in the same constricting algorithms as doing so on Spotify, there is no doubt that Tik Tok and Instagram impact listening. Record labels encourage artists to promote their music on Tik Tok, or make up dances to seek ever-elusive virality. But engaging with music on these platforms can also take a more traditional, word-of-mouth style. For 137,000 people on Instagram, Margeaux Labat (@marg.mp3) is a source of delightfully unfamiliar – and generally delightful – music recommendations. “Discovering new music is one of the great joys of my life, so the fact that I’ve played even the tiniest of parts in perhaps bringing that joy into other people’s lives is pretty amazing,” Labat told Mission.
“I discover a lot of music simply by being an active listener in my environment. I’m always that person that’s Shazaming in the corner at a restaurant, at a record store, at a coffee shop,” says Labat.
The recent popularity of ‘listening bars’, which first originated in 1950s Japan and now provide a social experience which foregrounds musical appreciation, also reflects that in-person approaches to music discovery have far from evaporated. Labat said she finds “a lot of great music through movie soundtracks and scores”. The resurgence of tracks like ‘Murder on the Dancefloor’ by Sophie Ellis Bextor – the 2001 hit featured at the end of Saltburn – or Kate Bush’s ‘Running up That Hill (A Deal With God)’, which reached number 3 on the Billboard 100 after appearing in Stranger Things, makes it clear that TV and film soundtracks, along with their repurposing on social media, have significant sway over young people’s listening.
“I find Spotify and those platforms to be amazing tools for music discovery. Spotify has over a decade of data on me and my music taste.”
Margeaux Labat
Labat’s pursuit of new songs mainly takes place “online, either through Instagram, Spotify, Youtube, Bandcamp and online radio like NTS Radio”, however. Though personal algorithms vary and many of her peers have different experiences, she says: “I find Spotify and those platforms to be amazing tools for music discovery. Spotify has over a decade of data on me and my music taste.” Though this can be “scary”, she admits, “what they recommend me is usually pretty great”.
A few days previously, McDonald told me he would prefer to see streaming services think “less about ‘personalization’ and more about building communities.” Labat’s online community epitomizes how sharing music discoveries can foster connection.
Instead of being stuck in an Espresso echo chamber, those who complain about Spotify’s recommendations can take more direct action to seek out the same benefits of music-sharing, and sourcing.
Homepage image courtesy of Sergi Kabrera/Unsplash.
