Art and commerce are unholy bedfellows, and nowhere more so than in the music industry — never a savory business, not in the mob days and not now. Liz Pelly’s Mood Machine, a dive into the history and workings of Spotify, is a useful guide to the way the streaming platform has altered the business of being an artist, but it’s one that avoids the conflicts between being a consumer and a patron of the arts.
Mood Machine is a combination of reportage, history, and analysis. Its great strengths are its interviews with Spotify employees, use of internal Slack messages, and bringing into English for the first time details from a number of early Swedish articles about the company. Spotify was initially designed as an advertising company; its paid tier was the result of a concession to major labels in early negotiations over streaming rights. This may come as something of a surprise to people who’ve listened to Daniel Ek, over the years, pitch Spotify’s founding as something done for love rather than money.
Spotify hit on the idea of playlists, a way of curating songs to be background music
The music industry in 2006 was desperate, having been kneecapped by piracy — first Napster, then a host of file-sharing applications. (Apple’s introduction of the iPod two years before the existence of the iTunes store arguably inflamed and worsened piracy.) Though Ek would occasionally publicly align himself with that history of piracy, he also hired Fred Davis — lawyer to Britney Spears, son of that Clive Davis — to introduce him to the labels.
Pelly really reaches her stride in her chronicle of “lean-back listening,” the end result of a cultural process that began long before streaming. Radio as background music had been common for decades. Muzak had infiltrated stores. Moby’s Play (1999) reached hit status through aggressive licensing, including in commercials. In its quest to grow past early adopters and music enthusiasts, Spotify took things a step further, diverging sharply from iTunes, and not just in its focus on subscriptions over one-time purchases. An iTunes listener was active, deliberately selecting what they wanted from a library of songs they’d (ideally) purchased. Instead, Spotify hit on the idea of playlists, a way of curating songs to be background music in life, just as they were in movies and commercials. Pelly quotes a former Spotify employee saying its main competition wasn’t iTunes, but silence.
Well, maybe. This is one of the places I find Pelly’s analysis lacking — she repeatedly quotes sources who claim people are afraid of silence without noting that silence is increasingly difficult for the average person to find. A white-collar worker no longer has an office with walls or even a cubicle — but they can drown out their too-close, too-loud coworkers with headphones that play inoffensive songs that won’t break their concentration. The noise of a coffee shop, a coworking space, or subway commute could similarly be obliterated, especially with noise-canceling headphones. A sleep playlist might be preferable to the sound of a neighbor’s newborn shrieking through the wall or the rager in the dorm room next door. But to recognize this, you’d need to spend more time thinking about the Spotify consumer.
This is maybe why “chill” became the word that was most important when it came to these playlists; the point was inoffensive background noise. Even mainstream music wasn’t immune: Pelly points specifically to Billie Eilish’s career as an example of how the chill playlist conquered pop.
The playlist model meant listeners didn’t have a relationship with artists
Playlists created totally different economic incentives than songs or albums. Spotify used small “feeder” playlists and then “graduated” hits to major playlists if enough people didn’t skip the songs. These playlists were exploited by major labels as marketing tools, and landing on one became desperately important to some smaller labels. Perhaps unsurprisingly, sleep playlists were particularly popular.
The playlist model meant listeners didn’t have a relationship with artists — they had a relationship with playlists. As a result, some labels stopped focusing on musicians’ careers and more on the playlist experience, Pelly writes, especially those concerned with “lofi beats.” This may explain why the percentage of older songs played on streaming services has been steadily rising since 2020; playlist-centric listening and label inattention make it harder for new music to compete with music people already know.
It was a short hop from there to one of the most compelling — and damning — parts of the book: ghost artists, also sometimes called fake artists. To feed its playlists, Spotify began commissioning “perfect fit content” (PFC) music that matched those playlists but was cheaper for Spotify. “By 2023, according to a review of charts and messages shared on the company Slack, over 100 official playlists were made almost entirely of PFC,” Pelly writes. From there, the even lower-rent option of AI-generated music started looking good.
And then there’s what looks like payola: Spotify’s Discovery Mode. In it, artists and labels accept royalty rates that are 30 percent lower in exchange for more promotion in Radio, Autoplay, and the assorted autogenerated mixes Spotify offers. Listeners aren’t told about the increased promotion. Discovery Mode has been profitable for Spotify, to the tune of 61.4 million euros in gross profit from May 2022 to May 2023, Pelly writes, citing an internal Spotify chart. The artists most likely to opt in were independent and DIY — those most desperate for promotion. This is just the beginning, of course. “In the long term (2025+ time scale), we envision between 30-50% of all recommendations on-platform will be influenced by discovery mode.”
“It’s not sustainable to put out challenging records.”
Spotify’s uneven playing field — one where major labels had more control than people realized, where Spotify was using cut-rate library tunes to compete with actual artists, and where most users weren’t, you know, listening — disempowered artists. The company’s response was to create a team where “my job is to make them feel like they can grow,” one employee said (emphasis mine). Spotify’s Creator team and its Spotify for Artists program are focused on getting artists to purchase advertisements on Spotify.
The pittance the platform offers artists is well known. While having a huge catalog of recorded music at my fingertips is wonderful for me as a consumer, it also means that niche artists such as Oneida are functionally competing for my attention not just with Beyoncé, but with Elvis and the Beatles. Darius Van Arman, a cofounder of the large independent label group Secretly Group, tells Pelly that Spotify means that monetization is based on what gets repeat listens. “It’s not sustainable to put out challenging records.” But Pelly doesn’t spend much time engaging with Spotify from a user’s point of view, and it’s one of the book’s weaknesses.
While I believe the value of art has nothing to do with sales, the recording business is, well, a business. The industry profits of the ’90s were driven in part by a format shift to CDs that required consumers to repurchase their old catalog, and also by price fixing. (Some of the businesses involved will look familiar.) And Van Arman’s remark reminded me of other shady practices in the bad old days when people bought albums based largely on guesswork. Labels sometimes put out misleading radio singles; in some cases, as with The Verve Pipe’s “The Freshmen,” the single version of a song wasn’t on the album at all. So a consumer spent $17 (more than $30 in today’s dollars) blindly, only to be disappointed.
How does Pelly characterize this? “People would buy the CDs at shows, or because they read a review, and even if they only listened to it once and put it on a shelf, it had value.” Well, it had value to the label, anyway.
Fraudsters getting paid dilutes the pool for real artists
Pelly is straightforward about her bias — in favor of indie labels and musicians — but she doesn’t make a strong case for why indie labels should matter. Sure, they are more likely to sign artists that make weird, sometimes difficult music. But I have a hard time making a full-throated case for any record label, partly because it’s not clear to me in this day and age they offer many advantages over DIY releases, especially when artists are expected to market themselves. Pelly is in a better position to explain why indie labels matter than I am, and I wish she had.
This bias against the major labels also leads her to dismiss streaming fraud. Spotify pays all rights holders out of the same pool — fraudsters getting paid dilutes the pool for real artists. The trickle of streaming fraud court cases suggests the amount lost by musicians is staggering; a guy named Michael Smith got $10 million out of streaming services with fraudulent plays, $60,000 of which came from Spotify. And that’s just one guy, engaged in just one flavor of fraud. Indie artists have had their royalties stolen by bad actors. And there’s AI slop that seeks to siphon listeners away from real artists.
In fact, fraud seems rampant. Eric Drott, an academic whose Streaming Music, Streaming Capital Pelly cites elsewhere, also writes about the widespread use of bots, click farms, spam, and soundalikes. Spotify’s playlists, which encourage inattentive listening, make it easier for fraudsters to engage in these tactics. And because playlists are so important, some scammers sell bulk streams to artists, making their songs more popular so they’ll land on important playlists. This particular style of fraud appears to target up-and-coming artists who are desperate to build an audience, but big artists have been accused of engaging in fraudulent tactics, too.
In 2021, Rolling Stone reported on a marketer who claimed he could get millions of streams for his clients. One method resembles payola — pay-for-play on certain playlists. This kind of fraud seems to disproportionately benefit major label artists: G-Eazy, who is named in the Rolling Stone article, was on RCA, a subsidiary of Sony. Another type of fraud surfaced when a number of Spotify users said their accounts were hacked to play French Montana’s “Writing on the Wall”; Montana was also signed to a Sony subsidiary. Streaming fraud is also among the allegations Drake has leveled against UMG, saying that bots inflated the number of streams on “Not Like Us.”
One way to deal with the big three’s outsized bargaining power is simply to break them up
Pelly is right that Spotify’s solution to fraud — demonetizing tracks with less than 1,000 streams — is bad, and she provides compelling evidence that major labels are the direct beneficiaries of the policy change. But that doesn’t make the fraud any less real, and it certainly doesn’t support Pelly’s claim that “the music industry’s focus on stamping out individual scammers can often seem like merely an attempt to distract from the systemic grift of it all.”
If music fraud is real, and there’s good reason to believe it is, indie artists are getting fucked twice: first by fraudsters, and then by Spotify’s “solution.” Some kind of content moderation might be better for artists. It would also be more expensive for Spotify. By demonetizing certain tracks, Spotify has placated the majors instead of implementing a solution that might have benefitted independent artists and saved itself some cash. This might have been a fertile area for Pelly to explore.
Because Pelly is such a firm critic of major labels’ sway over Spotify, I was somewhat surprised to discover not even a whisper of antitrust action in the policy section of her book. Music industry consolidation — which has been occurring at a rapid pace for the last 40 years — has been a major factor in making artists’ lives harder. Much of her book read as a compelling case for the breakup of Sony, UMG, and Warner Music, the big three labels that together “control 70 percent of the recorded music market,” Pelly writes. A service like Spotify offering approximately any song a listener can think of has to negotiate with those three, and they have extracted for themselves preferential treatment at the expense of smaller labels. One way to deal with the big three’s outsized bargaining power is simply to break them up.
What Pelly suggests instead is a combination of things: universal basic income, music co-ops, artist-run platforms, and library-owned streaming platforms centered on local music. (I suspect universal healthcare would also help; it’s usually more expensive to get health insurance as a freelancer than it is with a job.) She outlines a music labor movement and the Living Wage for Musicians Act — though its future under the next Trump administration looks grim.
Pelly is tiptoeing around saying the obvious thing: Spotify is a tremendous bargain for users
This would have been an ideal place to reckon with user behavior, suggesting, perhaps, a move away from lean-back listening and avoidance of Spotify’s various playlists. The collapse of music criticism and journalism in the last decade or so has made discovering new music harder for normal people and made playlists even more valuable. But all I can really make out is that Pelly’s embrace of nonstandard streaming platform styles hints that she would prefer listeners to not expect to have every song ever recorded immediately available.
At times, it feels like Pelly is tiptoeing around saying the obvious thing: Spotify is a tremendous bargain for users. I pay less per month for an enormous library than I did for a single CD in 1995. And while there’s a resurgence of interest in physical media, especially vinyl, accessing music that doesn’t take up a bunch of space in your house and that you don’t have to lug around when you move is also a win. So is being able to listen before you buy. It is easier than ever to be a music consumer, and the music industry’s profits have rebounded. But artists are still getting screwed.
I don’t expect Pelly to chastise the reader about that, if only because it risks alienating her audience. But it does highlight a specific weakness of the book: the vacuum-sealed approach to Spotify. What happens to musicians happens to all other types of creative people, just a little later. The final chapter of the book, proposing solutions, is weak in comparison to all that comes before it; a stronger version might have shown how what’s happening in the music industry is an echo of what’s going on in society more broadly. Art and its distribution are hopelessly tangled together.
That’s the tension, isn’t it? Music is both art and a commodity. The rise of streaming has pushed music even further toward commodity status, nudging people away from close, attentive listening. Mood Machine isn’t the final word on how this all happened, but the level of detail Pelly was able to get from current and former Spotify employees does explain a lot about how we got here.
As someone who loves music, obviously I’m going to pay for a cheap subscription to the majority of all recorded music — it would be crazy not to. But it’s also crazy to support the thing that dooms the people who create the music I enjoy. If there’s a way to cut this Gordian knot, Pelly hasn’t found it, but then, neither has anyone else.
