A subscription to the world’s most popular streaming service comes with artist exploitation, pseudoscience and the funnelling of profits into the war machine.
Who doesn’t know Spotify? It’s the most used streaming service in the world, coming in at 678 million monthly active users (the second is Netflix at over 300 million). You make playlists for parties and events; you hear others’ selections in cars, gyms, shops and homes; you put something on at your workplace. You go for a run or sit on public transport and listen to your favourite musicians—all on tap with the touch of a few buttons. A paid subscription from just A$13.99 a month guarantees no ads. ‘Every song in the world for less than your shitty airport meal. What could go wrong?’ This is the question music journalist Liz Pelly has been asking for years. In her debut non-fiction book, Mood Machine: The Rise of Spotify and the Costs of the Perfect Playlist, she digs deep into the inner workings of a platform that offers a seemingly utopian excess of choice yet belies much more sinister gambits.
As can be witnessed throughout history, most ‘novel’ ideas tend to be remixes of those from times past. Pelly begins Mood Machine by uncovering how Spotify was initially developed through the co-optation of the file-sharing website The Pirate Bay. Not long after Spotify introduced its freemium model in 2008, it launched a paid service to pander to major labels who wanted a share of profits. For all its game-changing potential, the mainstream music industry has largely dictated the way it operates. Pelly outlines this early in the book: ‘Spotify wanted to pay [the major labels] a percentage of advertising revenue, but [they] wanted to be paid per stream.’ Yet, even as the company advanced, its most unique offering—personalisation—can be traced back to predecessors that brought together pseudoscience and public relations to sell products.
Two precursors stand out. One is Thomas A Edison Inc., the famous inventor’s holding company for his many business pursuits. In 1919, its marketing arm sponsored research on the emotional impact of music, releasing a thirty-two-page booklet titled Mood Music featuring twelve playlist categories according to ‘what they will do for you’, such as ‘To Bring You Peace of Mind’, ‘To Stimulate and Enrich Your Imagination’ and ‘Stirring’. Customers could also hold ‘Mood Change Parties’ with their friends and fill out their own ‘Mood Change Chart’. Mood Music went on to guide how Edison Inc. would sell its home record players and signature Diamond Discs, a type of vinyl record.
Its most unique offering—personalisation—can be traced back to predecessors that brought together pseudoscience and public relations.
It was also around this time that US military general George Squier discovered a way to transmit sound through electrical wires rather than radio waves. These patents would be combined to create the music subscription service Muzak in 1934, which sold forgettable easy-listening tunes for businesses to pipe into waiting rooms and lobbies. In 1939, when William Benton—an adman who would become a US senator six years later—took over, the company led a new surge of dubious research linking music to workplace productivity and introduced Muzak into wartime industrial spaces. Pelly cites its influence on the WWII-era BBC radio program Music While You Work, which was created to boost workers’ moods in weapons and munitions manufacturing. Likewise, Spotify’s business model would capitalise on users’ emotions, all while having an insidious yet expanding role in the military-industrial complex.
When I told a friend I was reading Mood Machine, which I summarised as ‘about how Spotify sucks’, she raised her eyebrow. ‘Duh.’ She was totally right—of course Spotify is bad. Yet what is it about Spotify that needs to be brought to attention? Just how bad is Spotify, really? Is it like when Super Size Me and Fast Food Nation tried to expose McDonald’s for increasing profits at the expense of people’s health, including claims that it aggressively advertises to children and marginalised communities, exploits workers and initiates vexatious lawsuits against activists? Within capitalist realism—as Mark Fisher described it in his 2009 book of the same name, ‘a pervasive atmosphere conditioning not only the production of culture but also the regulation of work and education’—how long does it take for a mega-corporation’s existence to be so normalised over time that its human rights and ethics infringements become unremarkable?
Pelly’s book comes at a time when big business is increasingly being excoriated alongside horrific crimes against humanity. The Boycott, Divest and Sanctions movement has gained renewed traction during Israel’s US-backed genocide in Gaza, which has now seen the dropping of an estimated 70,000 tonnes of explosives, ‘the equivalent of six Hiroshimas’, on 360 square kilometres of land home to 2.1 million Palestinian people. (Even though McDonald’s overcame criticism in the 2000s, more recent links to Israel’s military have contributed to record global profit losses.) Less present in the media, the catastrophic humanitarian and political crisis in the Democratic Republic of Congo has resulted in the forced displacement of seven million people—a cataclysm that has been fuelled by the rapacious mineral mining needed for smartphones and so-called ‘green’ vehicles. Halfway into this necropolitical decade that began with a global pandemic, more and more people are becoming attuned to the untenability of the current system, especially as it stands stark against the libertarian-centric tech solutionist era of the 2010s that simultaneously ushered it in.
Nevertheless, journalists like Pelly are working hard to expose the fallacies of our time: what may look labyrinthine and complicated on the surface is obscurity by design, such that people feel ‘disempowered by the so-called expert class’ and resigned to their machinations. Part exposé, part investigation, the aptly titled Mood Machine throws light on a company most people associate with the kind of naive techno-optimism of the Mood Music study. Since Israel’s renewed attack on Palestine more than a year and a half ago, Spotify CEO Daniel Ek’s name has increasingly come up in relation to weapons manufacturing. In 2021, his investment company Prima Materia put €100 million into Helsing, a German AI defence tech company. Late last year, it was reported that Helsing had launched its first attack drone for use by Ukraine, signalling a significant expansion in the arms race that has been called AI’s ‘Oppenheimer moment’. Tangentially, Ek was reported to have hosted Donald Trump’s inauguration lunch in January this year, while also donating US$150,000 to the ceremony. And even though the funnelling of profits into empire and the war machine is unquestionably egregious, it’s really only several footnotes in an increasingly insidious story of harmful capital-first anti-human behaviour any way you look at it.
Spotify CEO Daniel Ek’s name has increasingly come up in relation to weapons manufacturing.
To outline the many dealings of an opaque and rapidly expanding twenty-year-old company is not an easy task. However, Pelly has been on the Spotify beat since the mid-2010s, perhaps a natural progression for an inquisitive mind already paying close attention to the way music is being consumed during a time of waning interest in analog media. On the book’s first page, she writes: ‘It’s a perspective informed by my background in DIY music spaces, where we try to live the reality that there are other ways to think about collective culture.’ From that vantage point, it’s plainer to see how art has been monetised and appropriated to suit the purposes of commercial industry.
Subscribe to stay in the loop
Be the first to read new stories, discover publishing opportunities, and access resources. For people who love great writing.
Mood Machine scrutinises Spotify from its founding as an advertising platform in Stockholm in 2005, a joint venture between Ek (then a bushy-tailed and bright-eyed chief technical officer at virtual dollmaking game company Stardoll) and fellow ad-tech executive Martin Lorentzon. Here, Pelly’s sources are wide-ranging, from advocates in the non-profit sector to musicians and indie label managers. Of the hundred-plus sources, about a third are de-identified former employees who share internal workings only staff can access, such as company Slack memos or insider understandings of how musicians are sorted into tiers based on the royalties they generate on the platform.
Throughout Mood Machine, Pelly provides an extensive chronology, though not a linear one, charting the platform’s rise to popularity via the rollout of particular features and PR campaigns. What’s noteworthy is that Pelly works to situate these changes in the social and technological landscape within history. Just as one can map Edison’s ‘Mood Change Chart’ and muzak’s invention onto the creation of Spotify, it’s also evident that what the company has put in motion over the years isn’t particularly trailblazing. For instance, the ‘Daylist’ has its origins in ‘dayparting’, a term from ‘broadcast radio programming that caters to certain demographics according to certain times of day’. Similarly, its marketing service, ‘Discovery Mode’, is akin to post-WWII radio payola, wherein record labels paid for their releases to be prioritised in radio queues. Elsewhere, there exist accusations that Spotify stole its annual Wrapped campaign from an uncredited intern. Alongside various cases of Spotify’s alleged ethical misconduct over the years, Pelly juxtaposes Ek’s media appearances that sell these ‘advancements’ as contributing to a kind of egalitarian haven for musicians, where it is often proselytised that ‘whatever succeeded on streaming was simply the will of the people’.
The company’s most well-known offence, of course, is how little artists are paid. At the time of writing, musicians reportedly make about $0.003–$0.005 per stream on Spotify, which amounts to $30 per 10,000 streams. (As a comparison, Tidal pays musicians the most at an estimated $0.01284 per stream. Another alternative, Bandcamp, which sells downloads and physical merchandise, takes a 10–15 per cent cut of record sales.) Pelly is particularly critical of Spotify’s pro-rata payouts, where earnings are calculated based on the number of streams you get, relocating revenue to big names at the expense of lesser-known artists. This is further eaten into by third-party providers that all artists are required to enlist in order to have their music on the app.
What happens for most artists in the end? You get next to nothing. And if you have less than a thousand streams in twelve months, then you won’t get paid at all—third parties be damned. While indie bands such as Vulfpeck have attempted to game Spotify by releasing a completely silent album and encouraging fans to play it while they sleep, they were eventually found out and banned from the platform, with any royalties forfeited. Meanwhile, pop star Justin Bieber, who has encouraged fans to act similarly in order to top the charts—in a system that already advantages his status—remains on Spotify. Throughout Mood Machine, Pelly does well to underscore how Spotify’s rapid ubiquity contributes to a landscape of overwhelming artistic homogeneity because it emphasises and prioritises a one-size-fits-all model that assumes independent artists would want to ‘conform to the norms of the winner-take-all popstar system’.
What happens for most artists in the end? You get next to nothing.
What is not as well-known, however, are the privacy breaches, data harvesting and algorithm bias, defined as the ‘dark patterns’ that are baked into much of social media. Even if using the app provides a sense of freedom of choice, Pelly’s findings are a damning look into what Spotify really is—a ‘surveillance apparatus driven by emotion profiling and pseudoscience’. All in all, one can say that Mood Machine is structured such that the facts build up and then the so-called reveal doesn’t dawn on you so much as it simply confirms what has been ringing in your ear all along. ‘If the company is capable of all the hyper-invasive practices it boasts about, that is obviously harmful,’ Pelly writes. But this isn’t the only concern: ‘If it sells advertisers on targeting it is not capable of, that also seems harmful, because it contributes to the normalisation of surveillance and deception.’
And so we return to muzak: music designed for anywhere and anyone, for nowhere and no one, to be piped into cafes, spas and start-up offices, just as it was once played in airports, elevators and hotels. ‘To Make You Joyous’, as Edison’s original Mood Music offered. ‘For More Energy!’ Cue the many jejune playlists blasting in a Pilates class. According to Pelly, ‘Spotify is obsessed with mood’, pointing out that the company has more recently applied for ‘emotional AI’ patents, indicating further investment in ‘mood data’—a market valued at ‘$20.26 billion in 2021—and is projected to triple that number by 2030’. But, predictably, opportunism trumps benefit. Tracks that populate the ‘lofi beats’, ‘pov: indie’, ‘Cocktail Jazz’ and other related genres—of which there are many—popularised on Spotify rack up millions of listens. Yet this type of music was internally devised and termed ‘perfect fit content’, introduced by in-house playlist editors as ‘a cost-saving initiative […] pushing that cheaper content when users went looking for certain moods’. To wit, one of Pelly’s most viral claims is that Spotify enlists ‘ghost artists’, musicians hired to fill playlists and minimise royalty costs with placeless background music (something that Spotify denies despite the testimony of informant session musicians in Mood Machine and elsewhere).
Certainly, while you may use Spotify to put together your own hyper-specific playlists, where Mood Machine succeeds is in debunking this particular strain of political apathy. Even if these ‘faux artists’ don’t actually exist, their vapid gibberish still ends up restructuring cultural production. This is to say that simply participating within Spotify’s universe means being inside the engine of its power. And although this is not at all unique to Spotify—only one part of a larger Big Tech-centric landscape that profiles and manipulates consumers (the Facebook–Cambridge Analytica scandal comes to mind)—when moods become influenced by capitalists via affective forms such as music, the inevitable result is that it shapes and reshapes how people experience art. It’s similar to Netflix: as Patrick Marlborough writes about in a 2022 Kill Your Darlings essay critical of the platform’s content, it ‘hacked at the way people think about and engage with TV and cinema, methodically whittling down people’s natural curiosity and inclination to explore old work and new, burying anything made before 1980 and doggedly chipping away at the idea of the “classic movie” until it was nothing more than some Gen-Xers’ idea of a teenage pizza party circa 1984’.
History reminds us that we have seen these patterns before: through the commercialisation of grunge and punk music, ideas and politics which might otherwise be considered unprofitable can be primed for co-optation—yet another form of artistic flattening. Via Mood Machine, Pelly argues that Spotify’s influence brings about the creation of music that will ‘be dictated by data biased by the influence of engagement and advertising imperatives’. One unsettling example is hyperpop, a genre which only gained visibility via Spotify’s editorial interventions—so much so that the uninformed may assume the platform coined the term—therefore ‘making a scene that was previously sprawling and complicated with sonically and geographically varied sub-scenes into something commodified and palatable’. This is a neat articulation of the ways data harvesting social media has contributed to the gentrification of the mind; we can map this across many art forms as we witness an increasing homogeneity due to artists replicating viral or algorithmically friendly creations in their attempt at visibility. Decontextualisation, too, continues to be endemic in these endeavours, as younger generations become divorced from their artistic predecessors.
And so we return to muzak: music designed for anywhere and anyone, for nowhere and no one, to be piped into cafes, spas and start-up offices.
The fact that Spotify has moved into podcasting and audiobooks in recent years is a clear nod towards the company’s aspirations—that they would one day like to act as a broker for many more forms of creative work in the same manner as Amazon, Netflix, et al. The publishing industry, perhaps one of the more antiquated industries in the world of artistic creation, has generally welcomed this move under the notion that the increased popularity of audiobooks on Spotify will help authors’ fledgling incomes. Yet there’s a lack of clarity around its compensation model, calculated from ‘which titles are listened to and how long they are listened to’. Publishers and distributors continue to take the lion’s share of what seems already meagre (sound familiar?), not to mention that Spotify listeners have admitted that they have bought fewer audiobooks since they were included on the platform. Once again, we reluctantly witness the myth of Big Tech-mediated decentralisation rearing its head: while accessibility does become foregrounded in these settings, the appearance of yet another third-party means that creators tend to be moved further down the food chain, and the devaluation of art continues as writers rush to replicate list-toppers and bestsellers. How many ‘the next Sally Rooneys’ have come and gone? Already, Spotify is ‘making it easier’ to release audiobooks narrated by AI.
In the streaming-consolidated era, it is the human imagination that is most at risk. Furthermore, if profits from our creations end up funnelled into places such as Helsing, then there is absolutely no net good at all—not only do artists become exploited and financially disenfranchised, war is literally funded through our presence on their platforms. At the time of writing, calls to boycott Spotify have become louder in the public sphere as it has been recently unveiled that Ek has increased his Helsing investment, saying that ‘AI, mass and autonomy are driving the new battlefield’. This new deal brings Ek’s total contribution to €1.37 billion, with him now also at the helm of the company’s board as chairman. While Ukraine might seem a testing ground for the time being, it may very well set a gruesome precedent in this era of what scholar Sayak Valencia terms ‘gore capitalism’, in which ‘death has become the most profitable business in existence’.
Spotify is not inevitable: I’m an avid listener of music, play in bands and have never had a subscription. Instead, I listen to community radio, buy vinyl, CDs or cassettes and download MP3s. There are artist-driven alternatives such as Bandcamp, which further supports independent artists by waiving its revenue in full on ‘Bandcamp Friday’, and hobbyist coders and software developers who are actively creating external software that allow you to download your playlists straight onto your phone or computer. Pelly herself has been reporting on ‘library streaming’ since 2021, wherein music is stored in library collections, to be managed in the public interest with public funding. There may exist many more alternatives if the bulk of us get off the Spotify ship. As Pelly rightfully asserts: ‘Fairness and sustainability aren’t buttons that just get permanently turned on or off; they are values that are continuously created and destroyed, negotiated and fought for and defended as technologies and formats come and go.’ And like any form of resistance itself, it only takes a groundswell for the possibility of change. There are no solutions in Mood Machine, other than that it should galvanise readers to find ways to widen the conversation about streaming, and to have discussions about what it means to intervene in the devaluation of artistic labour.
In a 2001 essay, Dutch architect Rem Koolhaas, observing the decline in the art form, coined the term ‘junkspace’ as something that ‘cannot be remembered […] because it cannot be grasped’. We can say the same of Spotify, which too ‘pre-empts people’s sensations’ and ‘[knows] everything about you, except who you are’—a nifty trick in a world of identitarianism that foregrounds hypervisibility and the mobilisation of personality as a saleable good. And as we have seen through the invention of new technology, convenience tends to come at a cost, which, as data scientist Colin Horgan writes, overrides ‘abstract ideas like privacy, democracy or equality’. If you really listen to today’s muzak, soundtracked to the hum of drones in the distant background while feeding you the idea that you can’t do without it, it becomes clear that it’s time to opt out. The unsubscribe button is right there.
