Sometime during Winter 2014 someone posted the following question on Quora: why does “my Spotify radio sounds so repetitive? I feel I am getting a few artists repeated in my Spotify radio.” Well, the Finnish ‘infojunkie’ Heikki Hietala swiftly replied (some time later in March), that’s because “the radio functionality in Spotify is very crude.” At the time, Spotify Radio had been around for more than two years, yet users seemed somewhat dissapointed. Maybe Spotify will “come up with something soon”, Hietala remarked, as for now “it’s very annoying”. Hietala apparently has had the same experience of repetition of songs played on Spotify Radio, and instead recommended the music streaming service Pandora. According to him, the latter had more succesfully “chopped the music up into tiny pieces of metadata, and [Pandora] are able to deliver a truly mesmerising radion function due to the vast amount of information they have on the music” (Hietala 2014).
Quora is a so called ‘question-and-answer’ site. Questions are posted—and subsequently answered, edited and organized by the community of users on the same site. Quite a number of questions on Quora evolve arund tech—which is hardly surprising since the company is co-founded by two former Facebook employees and based in Mountain View, California (Google headquarters). Quora also seems to be a site frequented by tech employees themselves. Tech Lead at Spotify, Erik Bernhardsson, has for example published almost 30 posts, some adjacent to the discussions around Spotify Radio. A couple of months after Hietala’s post, a similar disapproval reappered on Quora. In fact, almost identical questions around the dodgy functionality of Spotify Radio kept being repeatedly posted: “How do I get Spotify to stop playing the same few songs for every artists?”; “How do I teach a Spotify radio station to play a wider array of songs?”; “Why does my Spotify Radio play the same artists over and over for me?” (our italics).
The last of these question was asked by web designer, Bas Leijder Havenstroom, who in a re-entry specified in more detail what he was puzzled about: “I re-asked this one because this frustrates me as well. Even if I start a radio station based on a playlist with many many artists, I find that some (specific) artists keep coming back. I have the feeling that this all has to do with commercial reasons. I believe record labels pay Spotify to have their artists to show up in radio stations and random functions more often” (Leijder Havenstroom 2015). Apparently, the algorithms running Spotify Radio are identical, independent whether one uses the Free or the Premium service (the only difference is that advertisements play in the latter version, which also cannot stream higher audio qualities). The issue, however, if track repetition on Spotify Radio has commercial reasons remains obscure. Basically, the same uncertainty und unpredicatbility goes for trying to research the different algorithms regulating music recommendations on Spotify Radio. In addition, these vary and have naturally been altered and improved since the release of the radio functionality. 2012 was the year when Spotify began updating its desktop software with several new features, including a Pandora-like radio station. “Spotify to Take On Pandora With Radio service”, was subsequently announced online (Hachman 2012). The commercial whiff was hard to hide; an online radio offering “would advance Spotify’s strategy of attracting users with free, ad-supported services who can be converted later into paying subscribers”, Bloomberg reported (Fixmer 2012).
Today, Spotify Radio is one the standard features of the music service, available on all platforms. The Spotify Radio “lets you sit back and listen to music you love. The more you personalize the stations to match your tastes the better they get”, the company announces online (Spotify Radio 2016). Spotify Radio is arguaby a popular service. The functionality allows people (or rather various algorithms) to discover new music within the vast back-catalogue of Spotify, offering a potential infinite avenue of discovery. Then again, following the different conversations on Quora and elsewhere on the Spotify comminity Web for example, the service has also been disliked. In fact, it has repeatedly aroused disappointment, and even substantial criticism. This was definitively the case when the service was launched in 2012, or as user stated at the comminity Web: “better radio algorithms … there are too many repetitions” (lehwark 2012). Yet, as is evident from the later discussions on Quora and the Spotify comminity Web, the sometimes quite devastating critique has remained: “the terrible radio algorithm repeats the same songs over and over (see [the linked] thread, which has been going for 2+ years)”, user ‘tellure’ groaned in late 2015. “Need to update the algorithims for Radio”, the user ‘zaliad’ lamented a couple of months later, “the repetitions are SAD at this point within 1 hour I can easily hear the same song three times” (zaliad 2016).
Within our research project on Spotify we therefore decided to set up an experiment that would try to explore Spotify Radio, and ultimately the limitations and restrains found within ‘infinite archives’ of music streaming services. Our hypothesis was that many streaming services’ radio functions (like Spotify) appear to consist of a series of tracks that are played over and over again. Spotify Radio claims to be never-ending, yet there seems to be a loop pattern. If our hypotheses would prove to be true, what would such loop patterns look like? Are Spotify Radio’s music loops finite—or infinite (given that ‘the algorithm’ can after all choose between 30 million songs)? How many tracks (or steps) does a ‘normal’ loop consist of? Importantly, how is the size of a ‘music loop’ on Spotify Radio affected by user interaction in the form of skips, likes and dislikes? Does for example n amounts of ‘likes’ expand the music loop in terms of novel songs and artists? In short, how much preprogrammed imagination does streaming platforms really display?
In order to answer these research questions, at the digital humanities hub, Humlab (Umeå University), we set up an experiment in the form of a reversed engineered radio loop, with 40 bot ‘listeners’. Essentially, reverse engineering starts with the final and implemented product—in our case the Spotify Radio application within the streaming service desktop client—and takes it apart “seeking clues as to why it was put together in the way it was and how it fits into an overall architecture” (Gheel 2014:10). Our bots were Spotify Free users—with literally no track record; they had ‘heard’ no music before they were put into action. They were programmed to start Spotify Radio based on Abba’s “Dancing Queen”, document all subsequent tracks played in the loop, and (inter)act within the Spotify Web client as an ‘obedient’ listener, a ‘skipper’, a ‘liker’ and a ‘disliker’. On the one hand this article recounts, analyses and discusses the intervention we set up. Yet, on the other hand, the article also describes the background and the establishment of the ‘radio functionality’ at streaming services, as well as tries to ponder and reflect around the media theoretical concept of looping per se.