Spotify sent out its annual “Wrapped” playlists to users this week: flashy, personalized summaries of a year’s worth of listening. It’s the one day a year when we get a packaged glimpse of the streaming giant’s inner workings, when we’re reminded that, yes, Spotify has meticulously tracked each song we listened to, embarrassing or not, and it has the timestamps and hot pink graphics to prove it. Looking at it gives you the sheepish, almost giddy feeling of being caught in a lie.
Given Spotify’s formidable trove of data on global listening patterns, it’s odd how little of it we actually see. The company owns the Echo Nest, for instance, an offspring of the MIT Media Lab that trades in music analytics. Among other things, the Echo Nest has created an intricate map of 5,000 genres on Spotify called Every Noise At Once, informed by localized listening trends; here, “Indonesian shoegaze” sits clustered with “Viking metal” and “hard glam,” and oddball genres like “whale song,” “warm drone,” and “cosmic uplifting trance” float out on the margins.
Yet the genres that Every Noise has excised from Spotify’s billions of tracks are nowhere to be found on the platform’s actual interface. Browse through Spotify’s in-house curated genres and you find, in place of Every Noise’s 164 varieties of hip hop, there are only a handful of playlists, some called simply “Taste” and “teardrop”; under indie, a playlist called “Lorem,” a reference to incoherent “lorem ipsum” placeholder text. “The Joe Rogan Experience,” though, presently has its own widget, lodged between R&B and Rock, under Spotify’s “genres and moods” tab. (Which is troubling, but also raises the question: is The Joe Rogan Experience considered a genre or a mood?)
That Lorem playlist, which has nearly a million followers, features tracks from Abbey Road alongside SZA, alongside two-minute-long songs clearly primed for TikTok virality, all hooks and feverish drums. Lorem’s curator at Spotify, Laura Szabo, explained the absurd title this way: “It doesn't really mean anything at all,” she told Complex, “so the users can dictate what this sounds like.” But how can they, exactly?
This week, Spotify launched—to collective groans—a “stories” feature, à la Instagram and Snapchat. It’s still in beta testing, and the whole idea is pretty comical, but the feature is a logical progression for a company intent on treating music as a passive, individualized user experience, untethered from any collective reality. Spotify’s ethos in this way is much the same as Instagram’s, spinning user engagement into a bizarre sort of branding. It is evident in Spotify’s real-time friend “activity feed,” its showy Wrapped display, its hazy playlist titles.
Spotify has long since mastered the “genre-less,” emotive playlist, claiming platitudes like “culture” and “mood” as the basis for its curation. But in reality, the whole platform is genre-less. Artists are not labelled by their genres (though the Every Noise map already sifts through Spotify’s data to determine, say, the biggest names in Peruvian punk); they’re not sorted based on their place of origin, or even their ties to other artists or communities. Despite loud demands by artists, the platform does not include credits for producers, audio engineers, or other musicians on tracks. Tight-knit local music scenes, always foundational to genre and innovation, dissolve on Spotify. Discovery is severed from the geography of music, and channeled strictly through the company’s own algorithms.
“These algorithms,” as Earl Sweatshirt said, “are weird and undefeatable.” Much has been made of them: they can make deep cuts into hits on a whim; they can steer what we listen to, and how we listen to it. It’s difficult to opt out: Every month or so, I disable Spotify’s “autoplay” feature, which plays “similar” songs when a playlist or album ends, ensuring seamless, iterating music. Without fail, the feature re-enables itself.
Spotify’s automated listening aspires only to replicate a nebulous “sound” for users. This is its conception of music. Artists, facing an industry dependent on streaming numbers, are left to game an impenetrable system. Shady companies offer desperate artists Spotify playlist placement for hundreds of dollars per song; dozens of YouTube videos walk through elaborate strategies to optimize a track’s performance in curated playlists. This has been the case for years, but in recent weeks, perhaps emboldened by a music industry forced online, Spotify has made its playlist payola far more explicit.
At the beginning of November, Spotify launched a feature called “discovery mode” for artists. It’s a tantalizing offer: Opt in, and Spotify will “boost” your song’s standing in its algorithms—increasing the likelihood it will be heard—if you agree to accept a lower pay rate for boosted streams. The whole concept might feel less slimy if Spotify paid artists anything even nearing reasonable royalties, but it doesn’t. Bands with tens of thousands of listeners net less than $2,000 a year from their Spotify payouts. My most-listened-to song of 2020 made the band about thirty cents, per my calculations (and their label probably kept much of that). But for Spotify, my one hundred listens to some sad Sandy Alex G song was celebratory in and of itself.
“Discovery mode” only bodes poorly for artists. Playlists and automated discovery are critical to the success of new releases on Spotify, so there will be intense pressure to buy into the reduced streaming rates for new tracks. But it’s self-defeating: the more artists that use the feature, the less effective it will presumably be for any given song, likely worsening the already-pitiful wages artists receive. And raising Spotify’s revenues.
We shouldn’t pretend like any of this is new. Radio playlists, from their inception, reflected the wishes (and cash) of major labels over any kind of “organic” measure of popularity; such trends continued even when the technical practice of payola was banned in congressional hearings in 1959. (In his fascinating study of FreeMix Radio, scholar Jared Ball argues that the payola ban really only served to shift payments from DJs to wealthy station owners, strengthening their control of the radio’s rotation.) What Spotify claims to be curated, personalized listening is really just the same old corporate music of the vinyl age.
It feels heightened now, though. Maybe because private equity firms are spending hundreds of millions to buy up all the rights to pop hits. But perhaps also because, as Wrapped halfheartedly reminds us each year, Spotify’s immense digital library and massive reach hold real potential for music. Glimmers of that can be seen on platforms like Bandcamp or even SoundCloud, which have centered artists in their designs, and spawned movements of digital, underground music as a result. As much as Spotify adopts lo-fi, bedroom-pop aesthetics, it has only ever proved hostile to independent artists. The platform already knows all our bad listening habits. It might as well put them to good use.
—Katya
virtual past
At the turn of the twentieth century, a new genre of self-help book seized the American cultural sphere—the popular music songwriting guide. Written by successful entrepreneurs of the music publishing industry, with titles like So You Want to Write a Hit Song?, Song Exploitation Marketing, and Dollars and Songs, these books promised an easy path for those looking to get into making music—and money.
A few decades earlier, such books would not have been possible, mostly because there was little money to be made in selling music itself. Concerts, lessons, minstrel shows, sure—these all had an admission fee—but the music they involved ceased to be profitable as soon as it was done playing.
But as the American affair between consumption and entertainment solidified into an eternal bond, music too was radically transformed from a simple pastime, communal experience, and art form into an industry that has since ballooned into the behemoth we know today. The concept of “popular music” was an invention necessary to grease the wheels of this commercial growth. The popular song is as ubiquitous as the folk song without being in the public domain. It is music’s unique potential for collective experience, privatized.
Today’s pop genre gets its name not from some internal characteristic but from its ability to be widely disseminated, universally liked, and of course, easily sold. “Popular” is not a quality of sound—it is a quantitative measure of how many people have heard it. But the tireless efforts of early commercial songwriters and the burgeoning music industry they were part of made sure popularity became as integral to music-making as any other qualities of artistic virtuosity.
This was the most alluring promise of the songwriter’s how-to book—that popularity could be predestined, turned from a descriptor of an existing song into a formula for a new one. The ultimate goal of popularity was to minimize the specificity of music that tied it to a certain performance or people or place, making it both ephemeral and unprofitable. Popular music guides preached generality and admonished against anything that could make a song appeal to a certain community: “Your lyric must have to do with ideas, emotions, or objects known to everyone”; “Strive for the happy medium in thought and words so that both sexes will want to buy and sing your song.”
The invention of the phonograph, which turned into the gramophone and later, the record player, further enshrined the ability of music, and its publishers, to divorce themselves from the shackles of space and time. The sound of a song—not just its memory or its correspondent notes on paper—could now exist in many places, playing over and over again. And most importantly, this sound, and its profits, no longer needed to belong to the person who produced it, but rather to the person who midwifed it into the ears of its consumers.
Of course popularity, especially when subsumed under the mantle of profit, is hardly a goal that inspires beautiful art. Even the successful authors of songwriting self-help had no illusions about this. “The public buys songs not because it knows the song, but because it knows and likes the title idea,” suggests prolific songwriter Irving Berlin in his nine rules for writing popular songs. “Sacrifice lines you are proud of, even sacrifice rhyme and reason if necessary, in order to accentuate the title line effectively.”
“Songs must be simple,” he says in rule number eight. And finally, the last rule, perhaps most responsible for Berlin’s own wild success, and the one that I like to imagine adorns a wall of every Spotify executive’s office—“The song writer must look upon his work as a business.”
—Lizz
what we’re looking at
Object Lessons in Art in America Becca Rothfeld on Henry James and the ethical implications of love, beauty, and art.
No Rest for the Wiki in Bookforum Rebecca Panovka on the egalitarian ethos of Wikipedia.
A Is For Alienation in The Quietus Charles Ubaghs interviews Jason Lytle of the psych-rock band Grandaddy.
Purity and Security in Places Journal A history and cultural analysis of plexiglass.