What streaming does to musical work
By Nuné Nikoghosyan
University of Geneva
While streaming first appeared in the early 2000s and has become an everyday activity, a sociologist of music can’t help but question the reality of streaming services and their effects on musical work.
Services and prices differ by platform, but the principle is the same: online radios and streaming services such as Spotify, Rhapsody and Last.fm allow us to listen to whatever we like, whenever we like without having to own a hard copy. We can choose a song, an artist, a genre or any other type of category (such as “happy”, “female vocals” or “piano”) and end up with listening to “similar songs” or “similar artists”.
The merits of streaming have, for some time now, been repeatedly emphasized and even glorified. Audiences have cheap and legal access to virtually all the music they desire: for the price of one CD per month, we can now subscribe to a streaming service and listen to the equivalent of hundreds of CDs a month. Musicians, in turn, were to gain from the direct access to audiences that streaming gave them, all the while cutting production costs and bypassing the obstacles that intermediaries and gatekeepers represented. No longer would “being signed” depict the only to way to get noticed and succeed. No longer would getting noticed be correlated with financial investment. From a free market perspective, and in an idealistic democratic world of online radios, musicians were simply to find their audience by their talent and by the attractiveness of their creation.
In his essay “The Long Tail”, Chris Anderson (2012) explains this mechanism in detail while asserting that the future of the cultural industry lies in the potential brought by digitalisation. According to Anderson, online shops and streaming services – be it in music, film or publishing – have given rise to new markets and provided audiences with cultural products unavailable in traditional retail stores mostly due to lack of physical space and high costs of stocking. In the early 2000s, the online radio Rhapsody provided access to almost 20 times more music than Wal-Mart, whereas about 22% of the music found on Rhapsody was unavailable in any other retail store (see figure 1). The titles available online but not offline are what Anderson calls the “long tail” of the music industry. These lesser known songs still attract audiences and can therefore be lucrative, though less than the hits of the moment. Furthermore, some of these songs may even themselves become hits and success stories of the “long tail”, just as the band Arctic Monkeys famously used the Internet and file-sharing websites to promote itself before getting signed.
In other words, selling “less of more” is now slowly replacing the selling “more of less” business model prior to the Internet. The “long tail” prevails as a result of complex algorithms that remain commercial secrets, as well as folksonomy practices including tagging and categorization by listeners themselves. What has been bought and listened to by one user or what one’s “friends” on the website listen to or count as a “favourite artists” will generally be combined by how the artist or song had been categorized (genre, mood, type) by the website moderators and users in order to generate an automatic “similar artists” or “similar tracks” list. In fact, the audience plays an active role in determining the fate of the track on the “long tail”, though each individual at a microscopic scale. In this, the audience itself is hardly influenced by the music industry professionals (gatekeepers, agents, marketers or majors) and each listener’s taste is ideally shaped in accordance to those of others.
In sociological terms, this has come to be known as “disintermediation” (Miller 2014), or the practice of cutting out the middleman – the cultural intermediary – in shaping tastes and making supply and demand meet. According to Liz McFall (2014: 44), “The work to connect production and consumption in some way concerns almost everyone, almost all the time.” In other words, we all act as intermediaries, one way or another, and even more so on the Internet. According to McFall, the Internet has only emphasized this principle and allowed for tastes to be shaped directly by one another and with less intervention by marketers. However, online platforms, be they streaming services or shops, act themselves as intermediaries. What has changed, according to researchers such as Charles Umney, is the intermediary business model, for the online platforms have an enormous influence on both artists and audiences, largely tucked away in algorithms kept secret. For example, one of the most obvious effects of online agencies and booking websites providing a large database of musicians and bands for hire is the downward pressure on prices due to increased and direct competition on the platform.
In fact, the grimmer side of the story of online services is only now slowly emerging. While the use of online services is on the rise, data point at a continuing precariousness of workers, valid also for musicians on streaming services. The year 2017 was a turning point for the musical industry, marking the first time that revenue from streaming surpassed that of physical sales. In Switzerland, for instance, 27% of music was consumed via streaming in 2017, and with another 26% downloaded, more than half of all music bought during the year was dematerialized. In 2015, streaming counted for 20% of the market in Switzerland, and only 1% in 2012. However, Swiss artists are gaining less and less even if being listened to more frequently. Since users of streaming services pay a fixed amount in subscription instead of per song, the more they listen to, the more this sum is divided into smaller parts before reaching the artists as royalties. Artists are paid according to the total number of plays received in a month, by country. That means, the superstars pocket the most. The top 10% of artists dominate 99% of the streams on Spotify.
As the journalist Manuela Schnyder writes, the Swiss band Trummer (with about 750 monthly listeners and 510 followers on Spotify) received some CHF 108 for a total of 17000 plays via streaming, which amounts to about CHF 0.0064 per song. Having sold the same songs on an online shop such as iTunes, for about CHF 1.50 per song, 120 downloads would have amounted to the same income. According to Spotify, the average “per stream” fee paid out in royalties varies between CHF 0.006 and CHF 0.008 – the algorithm behind the calculations remains a secret.
Moreover, some of the already trivial income earned through streaming might still have to be paid to the band’s agent or label, depending on the contract terms. Some artists receive a mere 15-20% of their streaming revenue. In fact, traditional intermediaries such as agents do not disappear with the new online platforms and radios. The latter do not provide the (financial) means to completely bypass agents and marketers if one is to make a living out of his or her own music. For the artists on the “long tail” of the music industry what streaming services bring in added value is more symbolic than financial. Having been heard more times on a streaming platform may eventually lead to selling more CDs, downloads or tickets when performing live. With downloads and CD purchases on the decrease and, on the contrary, demand for live music on the rise (Frith 2007), the emerging picture is that of the least lucrative practice (streaming) being of service to the most profitable one (live). However, with most musicians who find themselves on the “long tail” of the music industry, this option is hardly a reality, simply because a rewarding live tour requires financial investment and a specialized intermediary, at the very least.
The precariousness of those on the “long tail” is explicit and continuously accentuated by the “uberisation” that streaming services underlie. Selling not songs, but plays, one at a time, each at a price of about CHF 0.006 brings symbolic renown on the Internet, but financially rewards the bigger fish. In fact, streaming services are designed to make listeners discover artists that are new to them – this is one of their selling points – which ultimately means that the platforms themselves contribute to splitting the revenue between a large and growing number of artists, only the most famous of whom have an almost guaranteed share considerable enough to be considered a true income. Earning less from streams than downloads or physical sales will ultimately push these lesser known or “ordinary” musicians (Perrenoud 2007) to having to diversify their activities even more, as it is more and more common for artists nowadays (Bureau & Shapiro 2009). We may even hypothesize that the streaming economy eventually leads to a reinforcement of the belief that culture is a symbolic good not to be “reduced” to monetary indices (Bourdieu 1971), for creation at all prices (literally) underlies the artist’s vocation more than being recognized as work (Perrenoud 2012, Heinich 2005, Sinigaglia 2017).
Since one of the main attractions for listeners on streaming services is discovering new artists, the lesser known creators have an incentive to make it to the “similar artists” list of a famous one who already has a high number of listeners. That means, not only being similar to a star, but also making easily recognizable music that can be smoothly categorized and tagged by a preferably large number of listeners who will use the same classification system (on the convergence of genres on the music market, see also Lena & Peterson 2008). The notion of replicating stars and the winning formula takes us back to Theodor Adorno who already in the 1940s criticized the selling of familiarity in culture and the standardization of songs, in terms of both lyrics and melody structure, amongst other cultural products. According to Adorno, “The schematic build-up dictates the way in which he [the listener] must listen while, at the same time, it makes any effort in listening unnecessary” (Adorno 1941: 22). Adorno’s notorious critique was by and large due to his fears of a standardized “mass culture”, aimed at entertaining “the masses” and diverting their attention away from politics and heavier issues, leading to a general paralysis of people’s critical abilities. Alas, the “listening more of the same” principle promoted by streaming services and their “similar artists” playlists bring us right back to Adorno’s mistrust of standardization. And this, amidst rumours of platforms such as Spotify even commissioning producers to make standard and easily recognizable music that will then appear in popular playlists, being distributed under different stage names – known as “fake artists”, having no other online profiles – for evading a share of royalties.
A part of the standardization of songs has recently been studied by the musicologist Hubert Léveillé Gauvin (2017). Having analysed 303 US top-10 singles from 1986 to 2015, his study found that the average 20-second instrumental intros of the 1980s have now become only about 5 seconds long, whereas the average tempo has increased by some 8% in the last 30 years, and song titles have become shorter. One of the reasons, he argues, is the “attention economy” or the short attention span of listeners on streaming services rushing to hit the “skip” button. The song has to immediately catch attention, as fast as possible, before being skipped. For Gauvin, all technological changes in history have somehow changed music-making and listening experiences, so this recent trend is more of a normal than a negative aspect of streaming. But, to take his argument further and to link it back to Adorno’s, what sometimes catches attention, especially in a “similar artists” economy, is not originality, but similarity that fits and even replicates the winning formula.
In the end, the so-called “democratic era” of the music industry has not, so far, proven to be as democratic as the Internet first promised. Celebrities are still the biggest winners in the streaming economy, whereas financial investment (coming form an agent or a label) is still a prerequisite for making one’s way out of the “long tail” (see also Bourreau et al. 2017). The lesser known songs gain perhaps more attention online and may potentially lead the artist to become more visible, but the ordinary musicians are still constrained and losing on many levels, and perhaps even more so with the spreading popularity of streaming. They need streaming services for the attention they may hope to find, minimal as it may be, whereas the direct competition online and the mechanism behind the redistribution of royalties is even furthering the polarization between the most and least famous, in both symbolic and financial terms.
Furthermore, apart from the effect of standardization of songs induced and accelerated by streaming, song-making as well as listening practices are also changing. In the 1950s, recorded music was mostly distributed as singles. Then, the album came about, to the point where “concept albums” became a notion and meant not just a collection of songs, but of a coherent work from beginning to end, almost a story (Tournès 2011). According to the music historian Ludovic Tournès, the Internet, via streaming and downloads, has shattered the dominant logic of music production, that of the album (ibid.). Moreover, the author argues that being able to buy the songs separately and juxtapose them while listening gains the listener more freedom in choosing what to listen to and when, bringing both listeners and musicians closer to the principle of the singles economy of the past. In fact, more and more artists are now regularly releasing singles more than albums, hence constantly staying in the centre of attention with new releases, while also featuring on other artists’ albums, much like on a playlist. Streaming services take this notion further, for the norm is for a song to be separated from the others on the same album while at the same time being assimilated to titles by other (similar) artists, by virtue of algorithms (see also Fanen 2017). Artists are selling less and less albums and songs, and more plays (“streams”) casually suggested by an algorithm. If all goes well, this should lead to selling more concert tickets – the more reliable source of income for ordinary musicians nowadays.
Adorno, Theodor (1941) “On Popular Music”, Studies in Philosophy and Social Science, Institute of Social Research, IX, 17-48.
Chris Anderson (2012) “The Long Tail”, in M. Mandiberg (ed) The Social Media Reader, New York University Press, 137-151.
Bourdieu, Pierre (1971) “Le marché des biens symboliques”, L’Année sociologique, troisième série, vol. 22, 49-126.
Bourreau, Marc; Romain Lestage; Françoi Moreau (2017) “E-commerce and the market structure of the recorded music industry”, Applied Economics Letters, vol. 24, n° 9, 598–601.
Bureau, Marie-Christine; Roberta Shapiro (2009) “Introduction : ‘Et à part ça, vous faites quoi?’”, in M-C. Bureau, M. Perrenoud, R. Shapiro (éds) L’artiste pluriel. Démultiplier l’activité pour vivre de son art, Presses Universitaires de Septentrion, 17-31.
Fanen, Sophian (2017) Boulevard du stream. Du mp3 à Deezer, la musique libérée, Le Castor Astral.
Frith, Simon (2007b) “Live Music Matters”, Scottish Music Review, vol. 1, n° 1, 1-17.
Heinich, Nathalie (2005) Etre artiste : les transformations du statut des peintres et de sculpteurs, Klincksieck.
Lena, Jennifer; Richard Peterson (2008) “Classification as Culture: Types and Trajectories of Music Genres”, American Sociological Review, vol. 7, n° 5, 697–718.
Léveillé Gauvin, Hubert (2017) “Drawing listener attention in popular music: Testing five musical features arising from the theory of attention economy”, Musicae Scientiae, vol. 22, n° 3, 291-304.
McFall, Liz (2014) “The Problem of Cultural Intermediaries in the Economy of Qualities”, in J. Smith Maguire, J. Matthews (eds) The Cultural Intermediaries Reader, Sage Publications, 42-51.
Miller, Toby (2014) “Cultural Work and Creative Industries”, in J. Smith Maguire, J. Matthews (eds) The Cultural Intermediaries Reader, Sage Publications, 25-33.
Perrenoud, Marc (2007) Les musicos. Enquête sur des musiciens ordinaires, La Découverte.
Perrenoud, Marc (2012) “Editorial. Entre l’art et le métier, l’émulsion symbolique”, Sociologie de l’art, Opus 21, 9-18.
Sinigaglia, Jérémy (2017) “La consécration qui ne vient pas. Réduction, ajustement et conversion des aspirations des artistes ordinaires du spectacle”, Biens symboliques, n° 1 [online]
Tournès, Ludovic (2011) Musique ! Du Phonographe au Mp3, Editions Autrement.
 “How Arctic Monkeys’ debut single changed the music industry and killed the NME”, The Guardian, 22.10.2015 www.theguardian.com/music/2015/oct/22/arctic-monkeys-debut-single-i-bet-you-look-good-dancefloor
 “New online live music agencies have oversized power over musicians”, LSE Business Review, 29.09.2016 https://blogs.lse.ac.uk/businessreview/2016/09/29/new-online-live-music-agencies-have-oversized-power-over-musicians/
 “Slipping discs: music streaming revenues of $6.6bn surpass CD sales”, The Guardian, 24.02.2018 www.theguardian.com/technology/2018/apr/24/music-streaming-revenues-overtake-cds-to-hit-66bn
 “La moitié de la musique achetée en Suisse est dématérialisée”, RTS, 27.11.2017 www.rts.ch/info/sciences-tech/9103991-la-moitie-de-la-musique-achetee-en-suisse-est-dematerialisee.html
 “Le streaming n’apporte rien financièrement aux artistes suisses”, Le Temps, 16.01.2017 www.letemps.ch/economie/streaming-napporte-rien-financierement-aux-artistes-suisses
 “Has 10 years of Spotify ruined music?”, The Guardian, 5.10.2018 www.theguardian.com/music/2018/oct/05/10-years-of-spotify-should-we-celebrate-or-despair
 “Spotify’s Year in Music shows just how little we pay artists for their music”, The Verge, 7.12.2015 www.theverge.com/2015/12/7/9861372/spotify-year-in-review-artist-payment-royalties
 “Are Spotify’s ‘fake artists’ any good?”, The Guardian, 13.07.2017 www.theguardian.com/technology/2017/jul/13/are-spotifys-fake-artists-any-good
 “’They could destroy the album’: how Spotify’s playlists have changed music for ever”, The Guardian, 17.08.2017