This is my playlist. There are many like it, but this one is mine A research on personalized Spotify playlists
This is my playlist. There are many like it, but this one is mine A research on personalized Spotify playlists
Leonieke Bolderman MA
Arts, Culture and Society Erasmus School of History, Culture and Communication Erasmus University
Master Thesis 8 June 2016
ACKNOWLEDGEMENTS I remember Dr. Berkers stating, during the master thesis class, that men were more likely to pick a thesis topic related to music. Luckily I did not disappoint this statistic.
But first and foremost I would like to thank my interviewees. I literally could not have created this thesis without them. An equal amount of gratitude goes to my supervisor Leonieke Bolderman, who I would categorize as ‘strict but fair’ in her comments. Furthermore, a big thanks to heavy metal music and alcoholic beverages. Both were always there for me in stressful times while writing this research. The same can be said for my classmates, or better, class ladies. Always there when I wanted to express my frustrations with the topic of my thesis. Of course there are other people who I should thank individually, but you know who you are.
TABLE OF CONTENTS CHAPTER 1 – INTRODUCTION ....................................................................................................... 1 1.2 Playlists ....................................................................................................................................... 2 1.3 Research Question .................................................................................................................... 2 1.4 Relevance ................................................................................................................................... 3 CHAPTER 2 - THEORETICAL FRAMEWORK ............................................................................... 5 2.1 Music and identity ...................................................................................................................... 5 2.2 Music in everyday life: meaning, behavior and playlists ...................................................... 7 2.3 Playlists and identity.................................................................................................................. 9 2.4 Music as a social tool .............................................................................................................. 10 2.5 Music meaning ......................................................................................................................... 11 2.6 Playlist management and the shift from physical to digital ............................................... 13 2.7 Music as enhancement tools ................................................................................................. 14 CHAPTER 3 - METHOD ................................................................................................................... 16 3.1 ‘Digital’ Elicitation ..................................................................................................................... 16 3.2 Interviews .................................................................................................................................. 17 3.4 Data ........................................................................................................................................... 19 3.5 Coding ....................................................................................................................................... 20 CHAPTER 4 – ANALYSIS ................................................................................................................ 21 4.1 Spotify usage............................................................................................................................ 21 4.2 Playlists and playlists maintenance ...................................................................................... 24 4.4 Spotify and identity .................................................................................................................. 30 4.5 Playlist and music sharing ...................................................................................................... 36 CHAPTER 5 – CONCLUSION ......................................................................................................... 39 5.1 Findings ..................................................................................................................................... 39 5.1.1 Complimentary to identity ................................................................................................... 39 5.1.2 Activity, emotion and gender .............................................................................................. 40 5.3 Limitations ................................................................................................................................. 42 5.4 Future research........................................................................................................................ 43 REFERENCES ................................................................................................................................... 44 APPENDIX 1 – INTERVIEW GUIDE............................................................................................... 47 APPENDIX 1.1 Interview introduction ......................................................................................... 47 APPENDIX 1.2 Interview questions ............................................................................................ 48 APPENDIX 2 – Transcriptions ..................................................................................................... 50
CHAPTER 1 – INTRODUCTION “Henry Rollins once said that music exists to put furniture in your mind, 'because life is so cruel and TV is so mean.” (1993, as cited in Frith, 1996, p.108). Music is all around us, whether it is background music in a café, the radio at work, study music, your favorite CD in the car, dancing on a Friday night in a club, or when listening to music via earplugs while doing groceries. In modern societies music can be listened to at almost any desired moment. Physical media such as CD’s are not even required for music consumption. Smartphones can hold an entire record store worth of music. But even more remarkable, we now have access to applications that allow us to consume a nearly infinite amount of music. One of such applications started 2008 in Europe. The Swedish based streaming service named Spotify. This service provides its users with an extensive library of music for a monthly fee of €9.99. It is even freely available if you don’t mind the occasional advertisement between songs. Given the fact that streaming music is a relatively new phenomenon, it has not received an extensive amount of research yet. Existing Spotify research is predominantly focused on legal issues regarding streaming music, technical aspects and amounts of Spotify usage (Ritala, 2013; Goldmann & Kreitz, 2011; Zhang, et al., 2013). All conducted from a quantitative approach. Not only has the widespread availability of online streaming services such as Spotify, Pandora, Last.fm and YouTube given the modern day consumer access to a tremendous library of music. It has also given users the opportunity to explore and expand their personal taste without spending additional money beyond a monthly fee. No user has access to more music than the other. Even a single new CD of a major artists is usually not available for such a competitive price. Spotify is also, if the user wants it to be, an open platform. Users can share their favorite music with others, whether that is via Spotify or via other social media such as Facebook or Twitter. However, to say that Spotify is a prime example of ultimate user agency regarding music consumption would be an oversimplification. Spotify is encoded with refined algorithms that can direct a listener into a particular direction by offering “suggestions” (Song, et al., 2012). Such tools are ‘Discover Weekly’, which feeds the user a lists of songs every week based on their listening behavior, or the option to explore ‘similar artists’, which is self-explanatory. Besides access to a vast amount of music and the ability to share, Spotify users can also create their own playlists. With a simple click of a button the user can add a particular song to a playlist and can categorize songs to their preferences. Given the fact that playlists can be created with relative ease and little effort, users have the ability to sort their music based on infinite criteria. Playlists can be created for a particular mood, for studying, for sport related activities, as a collection of favorite songs, for parties and so on. These playlists Page | 1
are a personal creation and are unique to each user. Finding a similar playlist containing the exact same songs seems as likely as shuffling a deck of cards in similar order two times in a row. Whether or not Spotify played a role in the creation process of these playlists, they are made for a reason by the user no matter how trivial this might be. But research on specific motivations for creating Spotify playlists does not exist as of now.
1.2 Playlists The most recent research, on music streaming services, has predominately been done from a market point of view and focusses on the changing landscape of the music industry in general (Aguair & Waldfogel, 2015; Adermon & Liang, 2014). This predominately ignores individual behavior and experiences. Alternatively the current research will offer a more in depth focus on the personal experience that music playlists evoke among listeners and how those experiences relate to the user his or her identity. Identity for this research is considered as how the created playlists fit into broader aspects of social life, for the interviewees. An example of a specific case would be dressing accordingly to the genre of music that is listened to. More broadly speaking would be how the playlists represent the personality of the interviewee, according to their own observation. As Firth (1996) states: “Music, like identity, is both performance and story, describes the social in the individual and the individual in the social, the mind in the body and the body in the mind” (p.109). Music is more than an expression of the listeners taste and a Spotify playlist would, along similar, lines not be just a representation of ‘greatest hits’ but an expression of the listener in broader terms. Existing research on how streaming services influence user behavior focused predominately on hours active on Spotify, what devices are used to listen to Spotify and listening session’s length (Zhang, et al. 2013). Changes in music reception due to streaming services have been observed by Avdeeff (2012): “It would be difficult to argue that digital music technologies and MP3 players, in particular, are the sole taste determinants, but they, nevertheless, play a part in the process” (p. 281). A streaming service such as Spotify can indeed influence consumer taste, focused from a quantitative perspective. However, the qualitative underlying process has not been studied and this where the current research can play an explorative role in, with understanding this playlist selection processes and how they are connected to the user on an individual level.
1.3 Research Question As playlists are created for various reasons music does not have to be only a tool of enjoyment, it can also spark emotions and memories. This research will look into the Page | 2
creation and maintenance process of playlists, and in what ways the music is in the playlists is meaningful to the streamer. The focus is therefore on playlists created or listened to via Spotify. The motivation for looking at Spotify is the immense popularity of the platform and the ability to create, edit and maintain your own personal playlists. Therefore the following research question will be explored in this research: How does the playlist creation process take place and what meanings do streamers associate with the music on their playlists? The idea of meaning is embedded into how the interviewees for this research associate their music (individual songs in the playlists) in the context of social life. This context could be associating playlists with certain moods, memories or activities.
1.4 Relevance Streaming services are more popular than ever and continue to grow1. Yet while their role in the way we consume music increases, research is lagging behind. With predominantly a quantitative focus, underlying processes have been ignored by contemporary research. This research will explore therefore relative uncharted territory. It is surprising how such a widespread phenomenon receives such little academic attention form the social sciences. However, considering the amount of time academic research and publications take, it is expected that this might receive much more attention in the coming years. There is also the factor that with new technology, such as Spotify, it is important to wait out the initial starting phase to see if it is something to stay or if it’s merely a hype. While this research, due to its limitations, cannot provide any generalizable conclusions regarding general consumer experiences, it can offer a new insight or at the very least a starting point for potential new research. Furthermore, research done by Avdeeff (2012) demonstrate that streaming can influence such matters as taste, but it does not account for how this process occurs. That is to say how it is experienced on an individual basis. Because this process is relatively unknown, this will be second part regarding the relevance of this research. Avdeeff (2012) does hint on the ‘how’ part of the question, stating that: “While it cannot be said that technology is fully accountable for musical taste or engagement, it does, however, exert its influence on how people listen to music, which ultimately affects their listening habits” (p.282). So rather than the streaming service itself playing a role regarding taste, that is to say Spotify influencing a user his or her taste. But because Spotify is being used it can affect the overall listening habits and with that trickle down to what is being listened to. Expecting
Statistia, Number of global monthly active Spotify users from July 2012 to January 2016. Retrieved from: http://www.statista.com/statistics/367739/spotify-global-mau/
Page | 3
Spotify to radically influence taste would be a step to far. But how music is being listened to, in this case Spotify, could potentially influence overall listening habits. At this point, Spotify is past the peak of being merely a hype. With rapidly increasing numbers of users2, streaming in general might become the definitive way to consume music. Because streaming diminishes the reliance on ownership of physical copies and makes music accessible at any time, it could influence our perception of music and how much we value it. By looking at the user created playlists, which is the only way to have some sort of ‘collection of music’ on a streaming service, it is possible to get an insight into what kind of value users attach to their playlists and music in general. Think of it as the playlist replacing CD’s as a sort of music collection. While the focus remains on playlists this research fits into the broader framework of how streaming services are experienced among its users.
Statistia, Number of global monthly active Spotify users from July 2012 to January 2016. Retrieved from: http://www.statista.com/statistics/367739/spotify-global-mau/
Page | 4
CHAPTER 2 - THEORETICAL FRAMEWORK The following chapter contains an overview of existing research specifically focused on music consumption, music meaning and how music is associated with identity. Within this theoretical framework there will be predominately relied on research that was either done prior to the existence of Spotify or research that does not directly discuss Spotify. The limited amount of direct research regarding Spotify made it therefore difficult to find direct matching research regarding the topic of research. Each of the seven sub-chapters that this theoretical framework is built upon, discusses a different topic that ultimately has a connection to the main research question. It starts with exploring the connection between music choices and our own associated identities; following up this sub-chapter the focus shift to how playlists fit in the broader aspect of music in everyday life. After this there will be a more in depth focus on Spotify playlists, how they are used socially and what meanings are associated to these playlists. The penultimate sub-chapter will look at so called playlist management, as it explores the shift from ‘physical’ media (i.e. CD’s) to the digitally created playlists. The theory chapter will be concluded with looking at music consumption to increase the performance of tasks such as studying. In so far as creating Spotify playlists for the sole purpose of being listened to while studying. The reason this sub-chapter has been added was due to the fact that six of the interviewees were students and it is was reasonable to suggest that they, at some point, had created a playlist to help with concentrating, or even enhancing their study capabilities.
2.1 Music and identity “According to DeNora (2000), individuals engage in a reflexive process of remembering and constructing their identities while listening to music, which can serve as a form of selfaffirmation and discovery” (Rentfrow, 2012, p.409). Music, on top of being a tool of enjoyment, can be constructive to our identity. Frith (1996) describes that music enables us to place ourselves in so called “imaginative cultural narratives”. We consume music with our body and mind. Not only do we judge the inherit quality of the music itself, but we also judge if it is good for us and if the music fits within our lifestyle. The judging aspect of music creates certain boundaries to what we decide as worthy to listen to or not, as Bryson (1996) demonstrated. While it might be difficult to put it into words, music is selected on the basis that it somehow connects to how we position our self in the social context, mostly related to what kind of music we as listeners exclude ourselves from (Bryson, 1996). This process might be done on more subconscious levels and does not have to be something that a music consumer necessarily pays attention to. But the music we do decide to listen to stems from an “emotional appeal” (Payne, 1980). Whether that is appreciation of quality, a connection to Page | 5
the lyrics, whether or not it is good music to dance to, or any other criteria. It is a personal affair, becoming part of our private sphere. As the music we have bought and listened to over the years is something that is uniquely ‘ours’. Ruud (2009) states that we like to think that we can somehow connect our music taste to our personality. “Music provides important peak or transcendental experiences which may strengthen the formation of identity in the sense that we feel meaning, purpose and significance in our life” (ibid, p.11). The role of music in our own identity formation is that it functions as a sort of soundtrack to our own life. The music we listen to is therefore not (or at least not exclusively) a representation of our identity, but something we use to construct our identity (Ruud, 2009). A platform such as Spotify provides potential opportunities for this identity building process. It has lowered the financial barrier, which existed much stronger when music was being bought on physical media, and gave access to a nearly infinite library. On top of that, there is the ability to create playlists that can be tailored and named in whatever way the user prefers. Likewise those playlists can be consumed in whatever fashion the user desires. Personal playlists can be enjoyed while studying or commuting to work. An automatically generated Spotify playlist with calm soothing songs can help you unwind after a stressful day. Or you can make a selection of up tempo songs to accompany you during a workout. With music being around more frequently it is not unlikely that how music is perceived in daily life is changing as well. “Bull (2005) argues that the iPod now offers the listener the unprecedented ability to continually readjust their mood whilst ‘on the move’, which previously was not possible with earlier music formats limited by their restricted mobility, and access to musical choices” (Lonsdale & North, 2011, p.109). In the same research Lonsdale & North acknowledge that most research done on the influence and importance of music listening is for a large part outdated. Few researches have been done recently. Which again is addressing a similar topic, considering the rapid growth a platform as Spotify experiences as well as an equally rapid changing internet landscape research is simply lacking behind. In June 2012 Spotify had 15 million users, three-and-a-half years later there were 100 million users3. While music can be associated with a wide range of emotions and memories it cannot express those emotions without context. For music to have an emotional bonding or connection to a particular event, there has to be something tangible. With that respect music is a facilitator of emotions that requires something outside the music itself. As Robinson (2005) put it, in order to experience an emotion of love in music there has to be a beloved
Statistia, Number of global monthly active Spotify users from July 2012 to January 2016. Retrieved from: http://www.statista.com/statistics/367739/spotify-global-mau/
Page | 6
person to connect it to. This makes music a tool to evoke a feeling that already exists. But music in and of itself cannot express a sought after emotion without context according to Hanslick (1986). “This consideration by itself suffices to show that music can only express the various accompanying adjectives and never the substantive, e.g., love itself” (Hanslick, 1986, p.9). Considering this, user created playlists on Spotify could in potential be an ideal way to facilitate such emotions. By grouping songs together into one playlist a specific memory or mood can be accessed at any time the user desires. Unlike an entire album of a single artist, where some songs might not facilitate such emotions. Because users are not required to buy entire albums, mood or memory related playlist could contain as many artists as the user likes.
2.2 Music in everyday life: meaning, behavior and playlists The selection process and the maintenance of a playlist is embedded in a larger framework of the overall music experience. That musical experience forms a major part in the daily lives of music consumers. On average most people listen to music more than they watch movies, television or read books (Song, et al. 2012). Besides the predominant factor of enjoyment, music can have other underlying significances. DeNora (1999) observed such an underlying meaning, drawing from in-depth interviews with women from various backgrounds and age categories. “Nearly all of these women were explicit about music's role as an ordering device at the 'personal' level, as a means for creating, enhancing, sustaining and changing subjective, cognitive, bodily and self-conceptual states” (p.34-35). The emotions associated with music occur at a personal level and are not explicitly interchangeable among users. This would speak for the case that a playlist indeed is a product that represents underlying significant meanings such as events and memories from the past. For instance, one of the interviewees from research done by DeNora (1999) stated that certain music reminded her of her father too much. This was considered a strong negative emotion for the interviewee. Drawing from this answer, music can therefore be purposely avoided for it can spark strong emotions that the listener find uncomfortable to be confronted with. Another possible example of strong negative emotions would be music played during the funeral of a close relative or friend. The music could purposely be avoided in future cases if the negative emotions the music spark are too strong. Yet relatively little previous research exists with the connection between music and (Western) funeral traditions. Exploring whether user created playlists contain songs with strong negative emotions (such as death) would fit in the connection between place and time. However this music does not necessarily have to end up in a user created playlist, but can be rather looked up when in a particular mood. For this
Page | 7
research music associated with negative memories will be observed as well and whether or not this music ends up in a created playlist. Reasons for listening to specific types of music are not done randomly, but are a selection process that is active and engaging. Beyond the emotional aspect as was observed by DeNora, in research done on the meaning of music in the lives of Australian elders the impact music has on self-identity has been observed extensively. “Participants used music as a symbol for defining their own sense of self and identity. Music is a symbolic representation of ‘who’ the participants are and how they would like to be perceived by others” (Hays & Minichiello, 2005, p.440). Selecting music in the case of Hays & Minichiello is not only done with a focus on music that the participants might enjoy, but also on what that music might say about themselves. Music as the representation of their own identity. This is an example that would speak in favor of listeners having agency on their selection process, or at the very least try to enforce agency over their own musical identity. The participants in this research were more aware of their emotions due to music and used it as a way to redefine self-identity (Hays & Minichiello, 2005). A different reason for music listening is provided by Sloboda, et al. (2001): “Behne (1997) has shown that adolescents reporting high frequency of personal problems also report higher frequencies of listening to music as some kind of escape” (p.12). In the same paragraph the researchers also state the fact that music can reinforce social identities (ibid). Which is something that especially occurs during adolescence according to Sloboda, et al. (2001). In similar fashion the idea of escapism can also occur among user created playlists. Think of ‘relaxation’ playlists, which can even be provided by Spotify itself. This does not necessarily have to be escapism in the sense of fleeing from personal problems, but can be escaping from stressful situations that can occur due to work or studying. In similar fashion but more focused on playlists in general, Cunningham, et al. (2006) observed that the creation of playlists, among respondents, is predominately related to emotional aspects such as mood and events. “A range of Moods were specified: for example, “feel good happy”, “mellow”, “aggressive, violent or angry”. Some mood descriptions are more complex and difficult to capture in words (or song)” (Cunningham, et al. 2006, p.241). There is a strong emphasis on emotion based playlist creation, with strong or outspoken emotions. Music that can make you feel happy, that can pump you up to make you ready for exercise, or in the case of a participant to enhance a particular feeling: “Ever just sit alone in the dark while it's raining out - you feel kind of lonely and sad, but it's that sweet, sensual sort of sorrow that just feels good and, in a way, comforting?” (ibid). Spotify allows users to change their music based on their mood with relative ease. Switching from an upbeat tempo when commuting to work or school to more slowly paced background music is done with relative ease. This makes a playlist distinctly different from a “mix” Page | 8
according to Cunningham, et al. (2006). For a mix similarities such as artists and genres are of more importance. The only drawback that could potentially limit the use of Spotify and the playlist creation/maintenance aspect on the go, is the requirement of a (fast) internet connection. Especially without the use of Wi-Fi, Spotify can drain the user’s internet data on, for instance, a smartphone. Spotify uses 96 kbs to 160 kbs bitrate for listening on a smartphone4. Which translates to respectively 0.72 megabytes per minute to 1.2 megabytes per minute. With daily usage this can add up quite rapidly and be a burden on the usually limited amount of internet data listeners have on their smartphones. However, the widespread availability of Wi-Fi should balance this problem out to some extent. There is also the ability to download songs from Spotify, so they can be listened to even when a device is not connected to the internet. This means that users are only limited by the amount of storage their smartphone, tablet or other device has.
2.3 Playlists and identity With Spotify providing an extensive library of music, users can create playlists that are a representation of their personal preferences regarding music as well as their personality and identity. Hagen (2015) has explored selection processes among frequent Spotify users and describes the idea of ‘context-sensitive’ playlists. What this means is that users create and listen to their playlists in order to enhance a certain mood, motivate themselves or bring back a memory of an important event (p. 12-13). DeNora (1999), made similar observations in her research on music’s social effects and stated that music can bring memories of past events to life. When a particular event is associated with a certain type of music or a song, hearing this music again can bring the memory back to live. “Music may thus be seen to serve as a container for the temporal structure of past circumstances” (DeNora, 199, p.49). Within this context a playlist can also function as a personal tool to recreate or relive memories associated to a particular event in time or a location. Other examples of place and time could for instance be: creating a specific playlist for a road trip or to enhance a certain mood that can be associated to a positive or negative memory. Existing research has already demonstrated that place and time matters in our reception of music. Gibson & Connell (2003) found that the personal connection created with the iconic Route 66 in the United States was predominately created via music. Visitors, especially from Europe, got familiar with Route 66 via the 1964 Rolling Stones cover of Bobby Troup’s song ‘Route 66’ (Gibson & Connell, 2003, p. 180). The exposure of Route 66
Spotify, General Questions. Retrieved from: https://support.spotify.com/us/learn-more/faq/#!/article/whatbitrate-does-spotify-use-for-streaming
Page | 9
within music can create a sense of longing for a place that the listeners have not yet visited. This demonstrates the imaginative power music can have for listeners with regards to place. It is not only about memory but also about the longing for an experience that the listener not yet had. Along similar lines, a user created playlist can be a well thought out process. A reflection of important memories that were of such significance that they had to be stored in a playlist.
2.4 Music as a social tool The previous sub-chapter focused predominately on individual experiences associated with music and playlists. But music, in its essence, is a “collective activity” (Håkansson, et al. 2007). Music recommendations form an extensive part of music as a social activity. Recommendations that are not based on streaming services’ algorithms but on relevant others or even strangers exists extensively likely due to this feeling of collective activity. Sharing music with others is not only done out of a willingness to share, but is “an essential way to expand the listener’s horizon” (Håkansson, et al. 2007, p. 311). The importance of sharing and recommending music to others becomes part of the selection process as well. Listeners do not rely on automatically generated suggestions per se, for in essence they are only a “suggestion”. Due to the social nature of music, a suggestion based purely on mathematics by a streaming service can be successful in finding similar sounding songs, but they cannot convey emotions or create a “collective feeling”. The collective nature of music would speak strongly in favor of recommendations by others over automatically generated recommendations. Besides being a collective product the initial consumption of music is usually done alone. Earplugs or listening to music at home creates a private listening sphere. Listening alone to music is a typical product of a modern (Western) society. Graves-Brown (2009) explains that “solitary listening” was deemed inappropriate during the early years of recorded music. “I fear that if I were discovered listening to the Fifth Symphony without a chaperon to guarantee my sanity, my friends would fall away with grievous shaking of the head” (Williams, 1923, p. 38). Music was considered a product that had to be experienced with the company of others in a specific location, whether that was a theater or concert hall. This is the complete opposite of modern music consumption where listening via earplugs can completely isolate someone from their surroundings. With that respect music is likely to have become a much more individual personalized product. Listeners might be reluctant to share their Spotify playlist, because it offers others an insight into their taste preferences and with that a look into their own identity. They might be ashamed of their “guilty pleasures” or don’t want to
Page | 10
share a song because it means something to them and this feeling cannot be passed on to others. A streaming platform such as Spotify makes it easy and more accessible to consume music alone, on your smart phone with earplugs for instance. With that regard playlists or songs in general could potentially be shared less than for instance a CD. Users are more than ever, or easily, able to bundle an individual selection of songs into a playlist. The playlists can be shared with followers, made public, or shared with a select group of friends. While it is reasonable to expect that users, to some extent, perform these practices, it is expected that most playlists are not intentionally shared due to the perceived private collection of music playlists might contain. Beyond social sharing and recommending Spotify itself uses refined algorithms to provide users with music recommendations. Discover Weekly automatically supplies the user with new songs every week, based on their listening behavior. It also looks at other Spotify users with similar taste as yourself according to Spotify5. This is an automatically generated playlist that still focusses, or pretends to focus, on the individual user. The playlist represents your listening behavior and can therefore be considered ‘your’ playlist. A representation of yourself via a computer generated algorithm without factoring in human perceived emotions. Recommendations done by streaming services could therefore be of significant less value. Finding music that connects to one’s identity can be of equal importance as listening to the music itself. A recommendation created by algorithms that focusses on non-emotional aspects such as rhythm and pitch cannot account for the personal connection that is created with listening to music. A possible reason for this could be that there is no gratification in acquiring music via automatic generated recommendations. It is handed to the user without having to work for it. Overall, it is a likely that recommendations by relevant others carry more value. A close friend or a relative can, to some extent, have an understanding of the listener’s emotions or feelings. Especially when music is a frequent topic of discussion
2.5 Music meaning Beyond the social aspect of music, individual experiences still form a major part of how music can be perceived or what meaning is associated to particular music. Songs associated with locations, like demonstrated by the song “Route 66”, have strong imaginative powers and can thrive fans to travel to particular locations around the world. But the other way around, where place is strongly associated with music or in particular artists,
Spotify, General Questions. Retrieved from: https://support.spotify.com/nl/learnmore/guides/#!/article/discover-weekly
Page | 11
occurs as well. Elvis Presley’s Graceland “has been legitimately compared with the shrines of medieval European saints, through the similarities in validation and devotion felt by the pilgrims” (Gibson & Connell, 2005, p. 202). The authors used the word “pilgrims” to describe the tourists and visitors of Elvis Presley’s grave. It is not merely a “trip” but a “pilgrimage” that every devout fan should do once in their life. Artist of a large enough magnitude can have a strong association with their “home” among fans. For instance, Bruce Springsteen and New Jersey, the Beatles and Liverpool, Bob Marley and Jamaica. Gibson & Connell (2005) discuss these artists in particular and their fan base as well. They are of such global fame that their background matters to their fans. But place in music can have a different significant meaning than associating the artist with his or her hometown. Music has the ability “to transport listeners away from their ordinary lives” (Gibson & Connell, 2002, p. 73). The sounds and lyrics of music are therefore not only instigators of memories but also of possible future events. A place you want go or live in the future. Much like the example of ‘Route 66’, songs are powerful imaginative tools that could inspire listeners to visit a particular location. This connection with place can be symbolic and does not have to be physical such as Route 66. For instance, the Beatles represent Liverpool in a more abstract way. They are connected to the city yet they did not record or create (for the most part) their music in Liverpool. “Indeed it seems ironic that the Beatles, who have come to symbolize Liverpool, wrote and recorded almost all of their music in West London” (Graves-Brown, 2010, p. 230). So linking the Beatles with Liverpool can be a symbolic connection, they are from Liverpool and therefore represent it regardless of the location where they actually record their music. While existing research focuses predominately on individual songs or artists, playlists can be the new form to create such individual experiences. Playlists made for a holiday could in potential achieve a similar feeling among its creator as some artists or songs would be considered a ‘fit’ with the location in question. Place in music is not as physical as a photo or painting representing a certain place visually. “Music, live or recorded, has a life, a dynamic which cannot be pinned down” (Graves-Brown, 2010, p. 238). Connecting music with locations or events is much more an emotional connection compared to something tangible or visible, but music can evoke a very personal emotion as well. A song that means nothing special for one listener, can have a significant impact on another listener. The song itself doesn’t have to discuss a place in particular, it can simply be a song that is heard somewhere which starts the process of associating the song with that particular location. Given this, strong emotion related playlists will less likely be shared with others via Spotify. They might even shielded of from Spotify followers. Spotify gives users the ability to make playlists ‘private’. What this means is that followers are not able to see these playlists. Page | 12
The importance of place and time in music is acknowledged, as they are the markers of important events in our lives (Ruud, 2009). Music that is listened to on holiday in the car with your parents, songs that are played at funerals, or more collective memories such as national anthems. Music can be in many cases a representation of a memory. Not only creating longing to visit a particular place, but as a reminder of something that we, no matter if it is a good or bad memory, do not want to forget. Music embodies the experience and can unintentionally trigger an emotion. A listener might not actively seek to relieve an emotion, but hearing music with the right rhythm, sound or lyrics can create a connection that starts the process of reliving or reminding a memory. While playlists are not necessarily used to experience such emotions, they are easy to use tools to group a selection of songs together based on emotions or mood. This provides users with a unique opportunity to control their music consumption. Creating a mixed selection of songs was, prior to streaming services, only possible by recording songs on tape via, for instance, the radio or downloading MP3’s and burning them on CD’s. This however did not change the fact that there could only be a limited amount of songs. Spotify playlists does not have such limitations, it is also easier to manage your music selection. Songs can be added or deleted as the user pleases. It is expected that user created playlists contain a diverse selection of songs that are added for reasons as to fit a particular mood or memory. While given the existing research these kind of playlists are to be expected, however playlists based on music preferences can exist just as likely.
2.6 Playlist management and the shift from physical to digital Formats such as CD’s or vinyl are, unlike their digital counterparts, physical goods. They can be stored, collected, touched, picked up and put on display to impress friends and family or as part of a collector’s hobby. This differs drastically from the modern Spotify user, who has access to a large amount of music at any given time but no user can “own” more music than the other. This does not mean that the playlist, created out of a nearly endless amount of music, has any less significant value. Hagen (2015) observed that heavy users of streaming services in Norway constantly maintain and change their playlists. The playlist can become a highly valued intangible good for the users. It is carefully looked after, altered and edited. It is something that represents a unique experience for every user. This would imply that the maintaining of a playlist is of equal importance as the playlist itself. Hagen (2015) acknowledged this: “Playlist management is as important as initial playlist creation, sometimes the purpose of the practice is its performance alone” (p. 20). Further examples that the music playlist is well thought process are provided by Cunningham, et al. (2006) in their research on playlists (and mix) creations, which has
Page | 13
already been briefly discussed. They observed that the initial creation of a playlist might come from a linear search from a personal collection, a list of songs alphabetically organized for instance. Yet after that initial phase is completed “the creator must imagine how each song might fit the theme of the mix or list: is the song of the same or a compatible genre as others already in the list” (Cunningham, et al. 2006, p. 244). While this describes a playlist creation process without using a streaming service, it demonstrates that the beginning of the creation process starts with an overview of what music the user owns and likes. After that has been mapped, the songs are put within different playlists that organized on different themes. These themes can be related to tempo, regarding lyrical content and even “emotional tenor” as Cunningham, et al. (2006) described it. The extensive maintenance of playlists and selection process of the playlist could therefore function as a modern day equivalent of the CD or vinyl collector. This would speak in favor of the fact that playlists are quintessentially a personal item. While they are not physical they can be discussed among friends and recommended to others. It functions as a way to demonstrate to others one’s music knowledge and the effort the user has put into it. However, the fact that a music playlist is a non-physical item has led to observable changes in the way music is consumed via digital media. Alonso and Keyson (2005) observed this even before the emergence of Spotify. They discussed the so called loss of the “physical experience” (Alonso & Keyson, 2005, p.1176). Acknowledging the fact that the object itself, whether that is a CD, vinyl or cassette tape, matters to the overall experience of music consumption. This so called ‘physical experience’ might be less of an issue for generations that are accustomed to using streaming services and other digital media at a relatively young age.
2.7 Music as enhancement tools Research on the influence of music in everyday life is abundant and can be found in the field of psychology, sociology and biology (Hargreaves & North, 1999; Peretz, 2006; DeNora, 1999). To conclude this chapter there will be looked at a different way music can be consumed. This type of consumption sparked predominately interest within the field of psychology. That is using music as an enhancement tool. In this case it is not concerned with mood or preferences per se, but more on what potential benefits music listening provides on performing certain tasks. Several studies have tried to demonstrate the potential benefits of listening to music and playing music. The “Mozart Effect” experienced a tremendous hype during the 1990s for the potential positive attributes listening to Mozart could have on performing spatial tasks (Hetland, 2000). While correlation has been demonstrated in several occasions, causality has been proven difficult to observe. Despite
Page | 14
the lack of causality, the potential benefits of music remain a subject of sociological and psychological studies. While the likelihood that music is actively used to increase cognitive skills by listeners remains open for debate. It does demonstrate that music, from the perspective of a researcher, is seen as a tool being used for more purposes than that of enjoyment. Similar to this research, while music is consumed for enjoyment there is also the element of what role music plays in daily life. Given this information it is possible that users create a personal Spotify playlist that is listened to when studying or performing a certain specific task. This would focus more on a concentration aspect rather than the enhancement of cognitive skills. Music with fast rhythms and power cords can distract listeners while calming and soothing tones can help to increase concentration. Kotsopoulou & Hallam (2010) explored this topic with various age groups in Japan, the United Kingdom, Greece and the United States and found that among students partaking in higher education, such as universities, music was used as a relaxation tool while studying. However, the authors provided a critical remark that the use of music as a study enhancement tool was of much less significance than was initially assumed. Listening to music while studying could be merely background music to break the silence. Done almost automatically with no literal intention to enhance concentration.
Page | 15
CHAPTER 3 - METHOD For this research a qualitative approach has been applied. This reason for this is that the research is not concerned with ‘why’ something is said but rather focusses on ‘what’ has been said by the interviewees. Becker (1996) described that: “Qualitative methods insist that we should not invent the viewpoint of the actor, and should only attribute to actors ideas about the world they actually hold, if we want to understand their actions, reasons, and motives” (p.60). Considering this research is not aiming to test a hypothesis, a quantitative method is less suitable, as that would shift more to ‘measuring’ data. That is to say measuring effects that Spotify has on individual consumer levels. In that respect this research extensively addresses the aspects of how playlists are created, or in other words what criteria are being applied, and what these playlists mean on an individual level. Due to the focus on meaning and playlist creation processes, semi structured indepth interviews are the most appropriate tools to observe such individual behavior. The observations give an insight into agency during the playlist selection process, or if playlists and music in general have any significant personal meaning. Also considering the fact that this research focused on frequent Spotify users that are actively creating and maintaining playlists, reliance on automatic generated recommendations was expected to be limited. Such a process can be done deliberately, where users simply ignore recommendations in favor of their own opinion. To get an understanding of creation processes and associated meaning to Spotify playlists, in-depth semi-structured interviews will allow the users to talk freely about their playlist creation and maintenance process. The in-depth interviews are, in that case, similar to a conversation (Legard, et all. 2003). Therefore each interview has an overall structure but differs per each interviewee, this forms the basis for every in-depth interview. “The first key feature of the in-depth interview is that it is intended to combine structure with flexibility” (Legard, et al., p. 141). This as well allowed some flexibility regarding certain answers provided by the interviewees, by going more in-depth on a particular topic when this considered a matter of interest for the interviewee.
3.1 ‘Digital’ Elicitation Because there is an emphasis on the creation aspect as well as how playlists are used, a ‘digital’ elicitation process formed the second part of the applied method. Prior to the interviews the interviewees were asked to bring with them a mobile device that allowed access to their own Spotify account. Most interviewees used their smartphones, while four interviewees used a laptop to access their account. The interviewees were asked to start up their account and open their playlists. Page | 16
Due to fact that it is difficult to talk about the playlist creation process, let alone the playlists themselves, this method formed the basis of the interviews. By engaging in a conversation while simultaneously looking at the playlists, the interviewees had a visual tool in front of them and did not have to rely solely on memory. This proved very helpful, as the interviewees immediately engaged in selecting the most important playlists they would prefer to talk about. Beyond the initial selection the interviewees on some occasions seized this moment as well to browse through all their playlist to see what they had created over the years. It made it a reflective moment for the interviewees as well, where they could see what they have been listening to and creating during the years they have been using Spotify. There were even moments of embarrassment expressed as users were confronted with playlists they have not listened to in several years.
3.2 Interviews The interviews contained some level of flexibility. As some interviewees provided shorter, to the point answers. This made reliance on follow up questions of greater importance. This also meant that the follow up questioned differed per interview as they were based on the situation at hand. However, for the majority the interviews relied on questions that have been created prior to the interview. These questions were always the same during each interview. The interview questions can be found in the Appendix 1. There has been an equal gender distribution with the ages of the interviewees varying from 20 to 45 (see table 1). All are active premium Spotify users (with the exception of one interviewee, who is not a premium user) who have created more than one personal playlist. The sole exception being an interviewee who does not consider himself a frequent Spotify user anymore. He switched to listening to radio out of discontent over the current way Spotify operates. In line with his discontent, this research is not intended to glorify the Spotify business model or the application itself. But users who have a strong favorable opinion of Spotify might be willing to overlook more of its potential flaws than others. All interviewees agreed to a face to face interview and approved the use of audio recording and the publishing of their first names. Any other names of persons that have been stated by the interviewees during the interviews are removed or anonymized. The interviews took place in a comfortable setting where the interviewees felt free to talk about their Spotify experiences and playlists. The length of the interviews varied, ranging from roughly 40 minutes to 70 minutes.
Page | 17
3.3 Interviewees Due to the large amount of Spotify users, finding suitable candidates initially appeared to be a simple task. However, many candidates that were approached stated that while they actively used Spotify their creation and usage of playlists was limited. Therefore potential candidates were asked if they had created at least more than one playlist and if they still maintained and created new personal playlists. This narrowed down potential interviewees significantly. After acquiring enough potential interview candidates, the selection process focused further on gender to ensure an equal distribution. Table 2 consists of graph representing the gender and average age distribution of Spotify users in the Netherlands. The interviewees are a successful representation of this table. By having a relatively wide age group, an observation would be possible on whether or not certain age groups valued their playlists and/or used Spotify in different fashion. Looking for interviewees was predominately done via inner circles. Face-to-face asking of friends and fellow students if they knew potential suitable candidates. This proved effective as the majority managed to point out at least one suitable candidate. Similarly a Facebook post also garnered plentiful response. After narrowing down the most suitable candidates, found via the Facebook post, two interviewees remained and were interviewed for this research. The remaining ten interviewees were al found via face-to-face request to friends and fellow students. All the interviewees have different backgrounds. Ranging from highly educated (master’s degree) to no formal degree. As table 1 already states, there is only one interviewee who is currently unemployed. Two of the interviewees lived outside Rotterdam (Amsterdam and Utrecht) the rest of the interviewees are all based in Rotterdam or in the suburbs of Rotterdam. There are three interviewees who do not have the Dutch nationality. Two of them are Swedish and one is Norwegian. All three of them do live and work and/or study in Rotterdam. Furthermore, as the researcher, I was already acquainted with four interviewees. The interviews were not done in any different fashion and the role of researcher-interviewee was maintained as properly as possible. Prior to the research I was not familiar with their Spotify behavior and music taste. With that respect it is also a possibility that the interviewees therefore might felt more comfortable providing answers about personal taste or what there playlists mean to them.
Page | 18
TABLE 1. Interviewees NAME Alex
Edgar Gerard Ilona Joep Jordh Laura Lovisa Maaike Nora Pelle Rosa
30 45 34 27 20 25 21 24 24 29 22
Male Male Female Male Male Female Female Female Female Male Female
PROFESSION Community Manager at an Art and Media technology center Financial Consultant Currently unemployed Senior HR Freelance Photographer Student HBO Student Master program Student Bachelor program Student Master program Student Master program Financial Account Manager Student Research Master program
TABLE 2. Spotify user’s age and gender distribution in the Netherlands 6
3.4 Data The most recent numbers provided by Spotify itself, state that the streaming service has 757 million users of which 20 million paying subscriptions worldwide8. With regards to the age distribution on Spotify in the United States, 62% of Spotify users is in the age range between
Retrieved from: Spotify For Brands While this research stated that 100 million users have been reached, Spotify itself has not released a press statement regarding these numbers yet. 8 Spotify, 20 Million Reasons to Say Thanks. Retrieved from: https://news.spotify.com/us/2015/06/10/20-million-reasons-to-say-thanks 7
Page | 19
13 and 29, with a peak among users in the age group 18 to 249. However, Spotify remains popular up to the age group 35 to 44 years10. This trend can be seen in table 2 when applied solely to the Netherlands. In similar fashion, this occurred as well when searching for potential interviewees. There was no outspoken dominance of certain age groups, but there was a slight advantage of users in their twenties over other age groups. While not initially selected upon, the interviewees do represent the average Spotify users, regarding age, in the Netherlands. For this research however, age played not a major factor considering there is no focus on such matters as the technological aspect, or a comparative analysis on the amount of Spotify usage. But significant differences related to the age factor, as has been stated earlier, will be part of the analysis if applicable.
3.5 Coding The interview transcriptions have been coded using Atlas.ti and were labeled accordingly to answer the main research question. By applying this method of analysis differences or similarities can be observed among the interviewees via a thematic approach. “The thematic approach is useful for theorizing across a number of cases – finding common thematic elements across research participants and the events they report” (Jupp, 2006, p.187). Because the interview questions were talked about abundantly in various ways by the interviewees, the coding had to be narrowed down in order to not become too detailed. That is why, for instance, negative and positive responses to playlist sharing have been grouped in the overall code of ‘collaborative/social playlists’. This was especially the case for codes related to ‘music meaning’ and ‘playlists’, as many diverse answers were provided the coding became to detailed. This has been purposely narrowed down in order to complete the analysis. This meant making a selection process based on what codes were necessary for providing a satisfactory answer for the research question. Due to the extensive nature of the interviewees many different topics were discussed. The codes have been initially grouped into families that relate to a frequently discussed topic. These topics, after grouping them into families, were put together in similar codes to have an overview of what has been discussed by the interviewees most frequently.
Statista, Distribution of active Spotify users in the United States as of April 2015, by age. Retrieved from: http://www.statista.com/statistics/475821/spotify-users-age-usa/ 10 Ibid
Page | 20
CHAPTER 4 – ANALYSIS “Music is like a pet. I can't have a dog in this apartment but music is my friend”.
Alex. “Spotify makes it just really easy. You become a bit spoiled, because I cannot live without Spotify. I had one month that my Spotify account didn’t work anymore because I changed from a paying account to getting it for free in my internet provider package. I choose deliberately for a subscription that included Spotify. Then I had to go month without Spotify, I thought it was a disaster”.
4.1 Spotify usage Two quotes about music and Spotify, provided by the interviewees. The role of music in their life and what a life without Spotify would look like. The two seem to go hand in hand in with each other. Talking about music with the interviewees it became clear that this a topic that everyone can share their own opinion on. Each of the interviewees expressed what kind of music works for them and what music doesn’t work for them. A very brief side note, heavy metal was the most disliked genre among the interviewees. Twenty years after Bryson’s (1996) research on symbolic exclusion and musical dislikes, is it perhaps time for an “Anything but Heavy Metal 2.0”? The interviewees used Spotify in varying degrees of amount but all of them had created a substantial amount of playlists. For all interviewees but one, Spotify was used frequently on a daily basis. Prior to the interviews one interviewee already stated that she had created two separate playlists. A “Rotterdam” playlist with music she discovered while living and working in Rotterdam and another one called “Oslo” with music that is associated with concepts as ‘home’. Both playlists are only used, listened to an edited when she is in one of the two cities. Here the lyrical content or song topics were of less significance. It is about associating music to a particular place. This was already in line with what Graves-Brown (2010) described with regards to locations associated with music. Joep stated that his usage has been limited because he usually does not know what he wants to listen to and prefers to be given be music via radio. For this interviewee Spotify is predominately used when he is looking for a specific (new) song. In general there were no significant differences in the amount of time that Spotify was used (on a daily basis) when considering gender and age groups. On average Spotify was used three hours per day, including maintaining and creating new playlists. The variables that influenced Spotify Page | 21
activity mostly were whether or not the interviewees were doing something else that limited there access to Spotify or listening to music in general. This was in most cases due to work related activities. Three interviewees explicitly stated that music listening was not possible during work. There were three interviewees who can be considered truly early adopters as they stated that they started using Spotify upon its release in the Netherlands or Sweden. Two of the interviewees claimed to have started using Spotify before its official launch of May 18th 2010 in the Netherlands11. These two interviewees had difficulty remembering the exact date they started using Spotify and it is likely that they overestimated the amount of time they have been using Spotify, considering it was not possible to access Spotify in the Netherlands before its official launch date, due to the fact that Spotify was accessible on invitation only at the time in Scandinavia, the United Kingdom, France and Spain12. The remaining nine interviewees started using Spotify between 2011 and 2014. Gerard was the latest adopter of Spotify. He expressed that had he many other tools to listen to music: “I am 45 now and I did 43 years without [Spotify]. It is an addition and it makes it [music listening] easier and more accessible. But without would not be a problem”.
The interviewee expressed that because he has been active with music decades before Spotify, or even the widespread availability internet, he possess many different possibilities to consume music. This is in stark contrast with Maaike and Laura who both stated that they cannot live without Spotify anymore. Ilona, also stated that it would be difficult to go without Spotify, but was more nuanced compared to Maaike and Laura. Alex was the only male interviewee who explicitly stated that he cannot live without Spotify. In general male interviewees were more indifferent or careless about the idea of not using Spotify anymore. Edgar expressed that although it will be difficult to quit Spotify, he has many different ways as well to listen to music, in similar fashion to Gerard, in that he can always rely on other platforms. Regarding activity and when the interviewees started using Spotify there was little difference between age and gender. Overall female interviewees did tend to rely more on Spotify as their only source of music listening compared to males. Female interviewees were more open and likely to talk about how music is important for them, how Spotify influenced their music experience and to what extent self-created playlists connect to their personality.
Spotify, Spotify Launches in The Netherlands. Retrieved from: https://press.spotify.com/uk/2010/05/18/spotify-launches-in-the-netherlands/ 12 The Guardian, Is Daniel Ek, Spotify Founder, going to save the music industry… or destroy it? Retrieved from: https://www.theguardian.com/technology/2013/nov/10/daniel-ek-spotify-streaming-music
Page | 22
All the interviewees were also able to express clearly what they considered the advantages and disadvantages regarding the usage of Spotify. The most notable advantages were the fact that Spotify offers a large library of music and is easily accessible. This was acknowledged by all interviewees. Other advantages were that Spotify was great for sharing music, it was fast and reliable and one interviewee also felt by paying for Spotify she is supporting the musicians as well. The advantages are consistent with the mission statement from Spotify13. Considering this, Spotify is doing exactly what it intends to do. The interviewees provided much more diverse answers regarding the disadvantages of Spotify. Ranging from the design of the user interface to the constant reliance on a steady internet connection. Two of the respondents also expressed that Spotify does not offer enough niche music and that it is not very easy to find new and upcoming artists. However, none of the interviewees felt like the disadvantages had any influence on the amount or way they used Spotify. Joep, who was the only respondent who used Spotify significantly less than others, expressed that Spotify lost its core strength by focusing more on major artists. This is similarly to the interviewees who stated that Spotify is somewhat lacking in any true niche music, or at least the ability to find niche music. While the advantages and disadvantages produced general consensus among the interviewees they also opened up a different topic among three interviewees, the feeling that music has become more fleeting due to Spotify. There is less sense of ‘ownership’ in comparison to a CD or vinyl for instance. As Alex stated: “I think that is also the thing with streaming. That both saving and throwing out music is so easy Like it's really just the click of a button, but going to buy a CD or throwing a CD in the trash… There was a connection there because you paid money, it's a physical thing and now music becomes more fleeting”.
Along similar lines Nora stated that she felt like music has become more ‘scattered’ due to Spotify. She feels less connected to the artists because it has become Spotify made it easier to switch to a different song or artist. This is due to the fact that all the music is available at any time. Therefore there is no incentive for her to feel connected to a single artist or band. These two interviewees were the only ones to express such feelings towards the influence Spotify has on the way music is experienced or consumed. Maaike made it clear that the music that she puts into her personal playlists is still ‘her music’. Even though it is a collection of songs available for everyone on a digital platform the idea of ownership still occurs. This feeling of ownership was observed among the majority of the interviewees,
Twitter, Spotify mission. Retrieved from: https://twitter.com/spotify/status/141929408763142144
Page | 23
especially when it regarded their playlists. Streaming is widely regarded as the very opposite of ownership141516. There is the very physical difference between owning a collection of CD’s and a collection of playlists on Spotify. A playlist cannot be held in your hands, it cannot be stored in a closet or put on display, but it still a collection of songs with which users seem to attach themselves to. In that case there is still a sense of ownership without holding a physical copy of the music itself.
4.2 Playlists and playlists maintenance The amount of playlists created did not significantly differ among male and female interviewees, nor different age groups. The most common types of playlists ranged from, party playlists, mood and memory related playlists, all-time and yearly favorites, activity playlists and study playlist. ‘Mood’ was a topic that was frequently discussed by the female interviewees and is strongly in line with the ‘context-sensitive’ playlists described by Hagen (2015). Overall the female interviewees for this research tended to use playlists more as a tool for a certain activity, predominately when they felt they needed to relax or calm down or on the other hand when they needed music for fitness related activities. So called workout playlists. For Rosa, aged 22, all playlists were associated with a particular mood. Her ‘autumn’ playlist consisted of music that made her think or experience that season via music. “I just really love autumn and the whole atmosphere, the cosines. It’s also a bit melancholic or something. I really love that. This is what the playlist ‘autumn’ reminds me of”.
She further expressed that playlists are a “snapshot of a specific moment”. Maaike provided an example of the purpose of a workout playlist: “Here I have the running playlists. This is just random top 40 songs that I usually don’t listen a lot to. But when I want to exercise and I want to listen to some happy music, the kind of music that easily makes you more stimulated [to exercise]. Then those are the kind of things I listen to”
Lovisa created playlists in similar fashion to Rosa but also had the tendency to connect playlist to individuals. She associated the music with her boyfriend, her boyfriend’s 14
CNN, Young listeners opting to stream, not own music. Retrieved from: http://edition.cnn.com/2012/06/15/tech/web/music-streaming/ 15 Business Insider UK, The one crucial reason Apple Music and Spotify can never replace your music collection. Retrieved from: http://uk.businessinsider.com/why-owning-music-beats-apple-music-and-spotify-streamingservices-2015-6 16 Yahoo Tech, Sorry, Apple Music I want to own my tunes, not rent them. Retrieved from: https://www.yahoo.com/tech/sorry-apple-i-want-to-own-my-tunes-not-rent-123410164954.html
Page | 24
sister and even created a playlist that reminded her of her best friend who passed away. Gerard was the only male interviewee who also had playlists connected to individuals. In similar fashion he created a playlist with music that was played at the funeral of his wife, who passed away in 2015. Both Lovisa and Gerard use such a playlist at the “right” moment. When there is something or someone triggering a memory of the person they had lost. These playlists will never be altered, they are fixed in time and only listened to when their mood and the situation allows it. If these playlists are put on, they purposely want to ‘feel’. “Only when I think of him. It's not like I am going to listen to this because there are good songs on it. But if I'm thinking of him and like missing him or something then it can be nice, like it's not nice in the way that I'm sitting there super happy but... It's still like yeah I don't know... A reminder.”
Lovisa. “Music is emotion, so that’s part of it as well. So I made a playlist with music that was played during the funeral … After the funeral I listened to it frequently. Now it’s still frequent on the moments that you miss someone.”
The music on their playlists remind them of someone they valued deeply. Contrasting this association with somber emotions, when the person itself represents a negative emotion, the experience would shift and more in line with DeNora (1999) where music is purposely avoided. This was not observed throughout any of the interviews. Of the remaining male interviewees none of them could reproduce such strong emotional connections to their playlists. In general the playlists created by male interviewees were more a representation of their favorite music, all grouped according to genre. Pelle predominately created playlists based on festivals he went to in the past and of his favorite genres. He expressed much less emotional attachment to the songs on his playlists. However, when asked he could reproduce specific memories he associated with the playlists. Such as a holiday with his friends or even a memory connected to a single song: “We always used to go to the same bar and the last song they played was always ‘Piano Man’, so now when we’re having a party at someone’s place we always put on ‘Piano Man’ when the party is almost over”.
Page | 25
These kind of memories are however not something that is specifically connected to a playlist or Spotify per se. Male interviewees did often connect playlists to specific memories, yet they never expressed any emotional attachment or connected it to a specific mood. The male interviewees could just as well have similar experiences but could simply be more reluctant to express this in an interview setting. While the reason for this is difficult to demonstrate, there has to be taken into account the fact that I, as the interviewer, am male as well. Perhaps a female interviewer would have garnered different responses, yet for this research such a comparative study is not possible. Functionality remained the driving factor for playlists created by the male interviewees. Edgar, created a ‘New York playlist’ with songs he associated with the city itself. This was done out of preparation for a holiday to New York City with his girlfriend. Although it is connected to a memory he is very fond of, it has no significant other meaning. He does not listen to the playlist to relive the experiences he had in New York City. But it does serve as a reminder of an important experience he had with his girlfriend. Nora was another interviewee that made playlists associated with cities. Namely, Oslo and Rotterdam. The reasons for these playlists were motivated by music she discovered while either living in that city or that she associated with “the mood of the city” in question. The songs on the playlists are connected to experiences, memories and do not have to relate to the city in the way that Edgar used it for his New York playlist. Nora and Edgar experienced in a different way how music can be connected to location. For Edgar there is the literal connection of songs and lyrics associated with New York City. In similar fashion to how Gibson & Connell (2003) describe music as being associated with location. For Nora the music did not necessarily had to relate in lyrical content to the cities. The music was associated based on personal interpretations and in that sense distinguishes from Gibson & Connell (2003). Yet music that not specifically addressed a particular location can still create feelings associated with locations considered important to Nora. In that sense there is a distinction between functionality, that is to say associating music based on lyrical content and music associated based on personal feelings. While the male and female interviewees seemed to differentiate quite substantially regarding what kind of attachment they had to their playlists, they agreed on one topic: there is no need to physically own music, streaming music is sufficient. This is a very stark contrast with research done in the United Kingdom among teenagers and adolescents regarding their music consumption habits. This research was conducted in 2009, before the widespread availability of Spotify in Europe, where the respondents expressed that “even if they were subscribing to an unlimited download service, 77% of respondents said they would continue to buy original albums” (Bahanovich & Collopy, 2009, p. 21). With the exception of Gerard, none of the respondents were still actively buying CD’s on a regular Page | 26
basis or expressed any urge to own physical copies of music. Again for Gerard age may play a significant role. Edgar, actively collected vinyl’s because he considers music to be his hobby. However, he does not have the same feeling for CD’s as he has for vinyl. The urge of physical ownership seems to have dwindled in a relatively short time. Bahanovich & Callopy (2009) state that: “Despite already having access to a variety of streaming services, 89% of all respondents stated that they still wanted to "own" music” (p. 22). None of the interviewees felt that they were experiencing a lack of ownership while using Spotify. This could be due to the fact that streaming services have integrated into music consumption in a way that users no longer experience an ‘ownership’ boundary. Spotify has experienced a dramatic increase in users17, the question is whether ownership still plays a significant role in the way users consume music. The rhetoric in research on dwindling sales in the music industry seems to be predominately focused on this distinction between owning music and streaming music (Premkumar, 2003, Bahanovich & Collopy, 2009). While the researches that have been referred to were conducted before the global rise of Spotify, they explain a scenario that has not been observed during the interviews. The interviewees did express some feeling of ownership over their own playlists. As Maaike provided an example of when her boxing trainer asked if he could use her phone to play some of her playlists via her Spotify account over the speakers: “But then I feel a sort of naked, because it is my personal music and there are songs on it that I really love that others might find weird”.
For her it felt like an invasion of her own “bubble of music”. While not necessarily touching upon the issue of ownership, it does demonstrate that personal playlists can still be experienced as being “my music”, even if it’s listened to on a streaming service. The playlist, if individually created, is a collection of songs that means something for the maker alone. While the interviewees like to share their playlists as well on several occasions, the playlists are predominately private affairs. Nora felt similar like Maaike and expressed that, although she is open for sharing, music is a more personal or at least individual affair. Ilona as well briefly stated that certain playlists were for private use only. None of the male interviewees expressed any similar feelings towards their playlists. Overall any other specific playlists related to themes as ‘workout’ and ‘relaxation’ (i.e. mood related) were not observed among the male interviewee’s. In similar fashion, male
Statistia, Number of global monthly active Spotify users from July 2012 to January 2016. Retrieved from: http://www.statista.com/statistics/367739/spotify-global-mau/
Page | 27
interviewees talked much less about emotional and feeling towards music. They predominately used rhetoric focused on the inherent quality of the music. But it did not diminish the fact that male interviewees were able to connect playlists, on multiple occasions, to a specific memory. The youngest interviewee Jordh connected one playlist to the first time he visited a techno festival. He was able to reproduce the exact date he visited the festival but again, the playlist was predominately a mixture of favorite techno songs that is constantly changing. The reasons why the female interviewees expressed more emotional attachment to music became more significant beyond what was anticipated. Research by Simon & Nath (2004) showed that the overall tendency in the United States is that women are viewed as more emotional and emotionally expressive. But whether or not behavior was consistent with believes remained unclear. According to their findings there is no significant difference between emotional expressions: “Overall, based on our findings from a national sample of adults, we conclude that there is little correspondence between men’s and women’s feelings and expressive behavior and gender-linked cultural beliefs about emotion” (Simon & Nath, 2004, p.1166). An older study done by Wells & Hakanen (1991) demonstrated that women tend to associate music more with emotions than men did. “Most respondents claimed to use music for emotional management, either mood enhancing (energizing, strengthening or lifting) or tranquilizing. Women engage in this management significantly more than men” (Wells & Hakanen, 1991, p.454). The results represent similar findings to that of the research done by Wells & Hakanen. Whether the male interviewees were honest about how they connect music to emotions remains unanswered. “Usually someone has like that question: "would you rather be blind or deaf?" Most people would usually say "I would never be blind, because than I could never see anything". To me I think, when I think about music and sound, that makes it difficult. Because to me it is really important. I think maybe it is because of the emotional thing. What I see can mean something, but the whole emotional thing that music adds… That would be a life without color. Yeah, it would make life quite dull”.
Getting back to the topic of memories, for both genders the playlists talked about during the interviews brought up specific memories connected to place and time. Similarly like Ruud (2009) demonstrated in his research, the songs on the interviewee’s playlists are markers of events that had a significant enough impact to be stored in a playlist. In many cases, especially among male interviewees, these memories had seemingly little emotional Page | 28
impact. Such as the memory of visiting a festival with friends, or a holiday. Much stronger connections to the playlists came predominately form the female interviewees. For instance, mood related playlists that reminded Rosa of a stressful period in her life. Another frequent reoccurring memory related theme were playlists consisting of music with which the interviewees grew up with. That is to say music they experienced via their parents. This has been observed and talked about extensively among at least five interviewees. Laura stated that she even shares a similar taste in music as her mother. Rosa and Nora had playlists containing music that their parents liked or recommended to them. Jordh in similar fashion has created a classic rock playlist so he can listen to it with his father when they are driving in the car. Among the interviewees who were still students, study playlists were an important part of their frequent listening behavior. Within the theory section the “Mozart Effect” has been briefly addressed. While the interviewees did not believe that listening to specific music enhanced their study capabilities per se, it did allow them to gain some more focus. It was also a very thought out process. The music had to be specific enough that it did not disturb the interviewees, but that it did allow them to focus on the task at hand. The music could not be too familiar otherwise the focus shifts towards the music itself, Maaike stated. Laura occasionally uses what she describes as “brainwave” music. Music solely focused on pitch that should help you to concentrate on specific tasks. According to Laura this was merely a “placebo effect”. Regardless, music was not put on randomly among any of the interviewees while studying. Maaike even described the importance of volume: “I tested it once and if I put the music too loud it doesn’t work. It has to be a certain volume and then it just disappears to the background and I can focus really well on what I am doing. Then I am no longer consciously aware of the music that I am listening to”.
Spotify has made creating such a selection of songs relatively easy, additionally it also provides similar automatic generated playlists as well. Among the interviewees still studying Jordh was the only who did not specifically create a study playlist. Unfortunately Jordh was also the only male interviewee who still studied and therefore any meaningful comparison is not possible. However, considering the other results it is likely that a similar trend would be observed. Overall, what Spotify does for the interviewees is strongly focused on the aspect of putting them into contact with new music. The interviewees experienced Spotify as something normal, a tool that has been integrated into their daily lives. When assessing the opinions of the interviewees, having such a large library of music available at any given moment doesn’t seem to be special or unique anymore, in a relative short time it has Page | 29
become part of the overall experience music. The interviewees are grateful for this and again, most of them could not live without it anymore.
4.4 Spotify and identity For Gerard, Spotify played little role other than making it easier to access and, if needed, share music. For him the age factor seem to be the most influential. His music taste has been shaped over several decades and Spotify played therefore a much smaller role. However, one of his first social interactions with his new girlfriend was done via Spotify. They exchanged playlists to get to know the other person. Spotify in such a way is complementary or even a helpful tool for social interactions. The strongest example of Spotify being complementary to lifestyle was provided by Alex who stated that: “I don't like owning stuff at all. In a way I was so happy with this music revolution. First with the MP3's. But even more when streaming came along. That I could get rid of my whole CD collection. All that plastic was just hanging around in my apartment”.
For Alex Spotify was the final push he needed to get rid of CD collection. For him this collection represents mostly wasteful plastic he no longer used anymore. Streaming is therefore the ideal solution, it requires no physical ownership. Something he has little interest in. For Alex there is no need for ‘physical’ ownership when it comes to music. These strong connections where Spotify would fit in a particular lifestyle have not been observed among other interviewees. For the rest of the interviewees it was predominately about the music itself. Spotify in that case plays a role as a facilitator. There was no real loss of the “physical experience” as was the case with Alonso and Keyson (2005). Albeit on smaller scale in comparison to Alex, the playlists created via Spotify by Jordh also represent a subculture that expresses itself via clothing. “I also wear this sling bag you know. I’m also wearing this hoodie now that says “patta” and that’s something typical for Amsterdam, you know like for a sneaker store”.
This is a style he associated strongly with hip hop subcultures. Spotify for him has broadened his horizon towards music. He experiences more music and discovers more new songs and artists. Therefore his playlists are more a reflection of the sort of music, or even sub culture, he wants to associate himself with. Yet he remained critical towards this and does not actively engage in this process on a conscious level. Also due to his relative young age he was, along with Lovisa, the only interviewee who already started using Spotify extensively in high school. For him it has already been part of his musical development when Page | 30
he was in his teenage years. When he started developing a taste for hip hop and techno music he was already familiar with Spotify and used it extensively. It is difficult to see what role Spotify did play in this process, considering the many outside factors that were also part of this process. Such as friends in school and/or puberty. Female interviewees were more likely to express that their music taste in general and the types of playlists on Spotify are a reflection of their personality. This is in line with the conclusions that DeNora (1999) drew in her research: “It [music] may be perceived to offer many types of semiotic particles that may be construed as 'emblems' for self-identity; it may provide a medium that comes to carry conventional or biographical associations” (p.53). The male interviewees were more likely to say that their music listening habits and playlists are not connected to their personality or lifestyle in general. Alex and Gerard, were the exception of this rule as they expressed the strongest emotional connections to music. Gerard simply states that: “Music is emotion”. Among female interviewees Rosa expressed strong connections to her music preferences and personality as well. The concept of the ‘underdog’ as she stated it. The music she listens to is what she considers, beautiful but at the same time ugly as well. It’s unpolished, raw, as a reflection of life in general. She is quite fond of this idea and feels like this fits well in her own perspective on life. But again Spotify is used as an accessory to something that already has been established. Laura expressed that she had difficulties explaining how music connects to her personality or identity. She stated that she couldn’t live without Spotify anymore and that it did change the way she experiences music, but was not able to go into great detail about how exactly it changed her music experience. Rather than being a tool that helps to construct identity, Maaike felt like she needed to protect her own (musical) identity via Spotify. “They [friends], or at least some of them listen to the same music as me but it’s usually very commercial. The stuff they listen to. On occasion I listen to that as well because it’s very catchy. That’s how your brain works. But it is not the music I will put in my playlists. Maybe I’ll do that on purpose to protect my own identity”.
Spotify is an open platform and the fact that someone can look at your playlists, if they choose to follow you on Spotify, can seem like an invasion of something that is considered to be private. She purposely made certain playlists private to protect her from this potential invasion. This an example of where Spotify is used as a medium to express or protect identity rather than a medium that shapes or contributes to the formation of identity. The results have shown that the biggest differentiator was the gender of the interviewees and to a lesser degree age. Interviewees of age 25 and bellow have fully Page | 31
integrated Spotify into their daily lives and cannot imagine living without Spotify or any other equivalent service. Interviewees in the age group between 30 and 35 were also fully adapted users and Spotify formed their absolute main mode of music consumption but they could potentially live without using Spotify. Above that age threshold users could still be accustomed to a live without streaming services, like Gerard, that Spotify could easily be replaced with others tools. The difference is that interviewees of the age group above 30, remembered vividly a time before streaming music and experienced more so to ‘gratitude’ towards Spotify, They felt like it was a tool that really helped to broaden their musical horizon. It opened new ways they never could experience before the existence of streaming services. Playlists connections with personality were most strongly expressed by the female interviewees, while some male interviewees stated that their playlists collection did not represent their identity in any significant way at all. As Edgar talked about what his playlists tell about him: “That I like music and that I keep myself occupied with it. I call it a hobby of mine. That’s what it says about me, but not much more I think. It is pretty diverse, I don’t only listen to very happy or very sad stuff. I don’t listen to a particular genre. No, it’s just my hobby and I think it’s really great”.
The playlists did not represent him in any significant way. Female interviewees were more likely to reflect their playlists onto their personality. As Rosa stated: “That I am very chaotic. It is really chaotic, there is no overlap or overall tone … It is based on feelings”.
Nora expressed similar feelings: “I think it says that I'm a very scattered person. Like a little bit all over the place.”
Coincidentally this sentiment was also experienced by Lovisa. When asked what her playlist say about her personally: “A lot. They would say that I'm a very messy person that tries to organize things. Like I am a very organized person because I'm a mess I think”.
Three times similar feelings were expressed, all done by female interviewees. Three induvial overarching sentiments, chaotic, scattered and messy. Male interviewees did not present Page | 32
any similar feelings at all. Gerard stated that the playlists don’t say anything about him personally but that his physical music collection does. He considers it something that is a constant work in progress that changes along with your musical preferences. The interviewees that started using Spotify at later stages in their life, usually after puberty or their teens, had at that point already established a firmly rooted musical taste. Yet how Spotify influences the way the interviewees experience music and how they are able to connect it to their identity has proven to be greater than was initially expected. Music is inherently a personal reflection for the interviewees and Spotify gave them a blank canvas, that is to they could start from scratch by creating and organizing their playlists, were they could experience music in their own and sometimes very personal way. As Ilona talked about what her playlists collections says about her: “Because you use it [playlists] as something personal. I think it is a sort of reflection what I am doing. Because I really listen to the lyrics. So when it is about the songs then there is a certain message behind it. I probably attach a lot of value to that message. So I think you can see what kind of mood I am in depending on what music I listen to”.
These playlists brought a larger framework to the table for the interviewees regarding music preferences associated with identity. This was not frequently reported among male interviewees. While among female interviewees such associations were occasionally observed. In one case an interviewee told how others can make judgement of their personality based on their choice of music. As Lovisa told: “I don't know what it is, but I think people associate music with who you are a lot. I think for me it's a bit of a mismatch. Once when I was in primary school I was like fourteen or fifteen. Friends of mine found out I listened to hard rock at the time and they were really surprised and like "oh I thought you would like ABBA and stuff" and I was like... I was really offended actually, "I look like an ABBA fan to you? So I think it's a bit of mismatch because people don't really think that way about me”.
Her friends attached a certain musical preference to Lovisa without actually knowing what she listens to. For Lovisa Spotify was also a reflection of her musical progression throughout the years. She felt embarrassed looking at some playlists, stating that she could not believe she used to listen to certain music in the past. Interviewees who used Spotify for a longer time, early adopters, had similar experiences as they talked about old and forgotten playlists consisting of music they used to listen to. Often the interviewees no longer listened to the type of music existing in their earliest created playlists. Their playlist collection is an overview
Page | 33
of their musical development. Reasons for not removing those playlists, range from ‘nostalgic’ feelings to simply not having thought about removing them. ‘Nostalgia’ is a term that occurred frequently throughout the interviews. This could be focused on the process of buying vinyl for Edgar, but was frequently used to describe certain playlists by the female interviewees. These nostalgic feelings focused predominately on childhood memories or music the interviewees listened to in the past.
“That it is when I'm in a nostalgic mood I guess. Going back to like childhood feelings. It is just kind of like with the guilty pleasures as well. You're a little bit "oooh...”, turning the sound down a little bit. But still really enjoying it. Kind of a more happy feeling about it”.
Music can have the ability to trigger nostalgic feelings, but in existing research this is predominately related to individual songs (Barrett, et al. 2010). But in the case of Nora, an entire playlist was devoted to ‘nostalgic’ music. The reason for accessing this nostalgic playlist, was when an outside factor triggered a particular feeling that made her want to listen to that playlist. Similar to the findings by Robinson (2005) and Hanslick (1986), the music in the nostalgia playlists requires and outside trigger in order to have an emotional (music) experience. Beyond emotional connections, the predominant influence Spotify seemed to have on the interviewees focused on accessibility and availability. Equally stated among male and female interviewees was that Spotify gave them access to new music they would not have looked for otherwise. Descriptions like “it broadened my music taste” were frequently expressed. For Nora Spotify made her less ‘conservative’ regarding her music choice. Lovisa experienced this as well but stated that there was a turn side to this. “I think maybe it makes you a bit lazy. Like when there was YouTube I think you had to be more specific and maybe you were more... Creative in your hunt for music. Spotify just feeds what they think is best”.
That ‘laziness’ comes from the fact that Spotify provides users plenty of music recommendations, either via services such as ‘similar artists’, the feature that looks for similar sounding bands or artists as the one you are listening to at the moment. But for the interviewees Discover Weekly was the main tool used to expand their own music taste. This weekly provided playlist, based on listening behavior, was used frequently by seven out of Page | 34
the twelve interviewees. Joep, the least frequent user of Spotify, found Discover Weekly a flaw in the Spotify design. According to him it didn’t feed him diversified niches, it still relied heavily on household names. Pelle, Edgar, Gerard and Rosa were the other interviewees who did not rely on the Discover Weekly function. Pelle even stated that he did not know exactly what Discover Weekly was. The other interviewees were positive about the tool and Alex explained that relying on Discover Weekly made him put less effort in his search for music, due to the fact that Spotify provides him music on a weekly basis. There was no significant difference between the ways this feature is used among male or female interviewees. Tools as Discover Weekly and ‘similar artists’, seemed to have greatly affected the variety of music the interviewees are exposed to on a regular basis. While they stick to their preferred genres, they relied much more on these services than was initially expected. These tools seem to be very effective at understanding the listeners taste and preferences. The limitations of, especially, Discover Weekly is that it usually provides recommendations that the interviewees were somewhat familiar with. They preferred to be fed newer ‘niche’ music. For Edgar and Rosa it is ultimately more important to find new music on your own and solely relying on automatic generated recommendations limits their overall music experience. But in general the opinion towards automatically generated recommendations was positive. This was not initially expected as it was anticipated that there had be a sense of fulfilment when looking for new music. The “collective activity” as Håkansson (2007) described it, regarding music, seemed to be of less importance to the female interviewees. For them private listening favored using music as a collective activity. For male interviewees the collective activity appeared much stronger. Alex valued the collective nature of music especially: “In the end of every year I try to make a list of like, ‘this is the best of this year’ and I listen to them a lot. I share them with a lot of friends. A few of my friends in Sweden, we do that every year. We make ‘these are my tracks of 2015’, 50 tracks or something, and then we switch that with each other, so we can what we have been listening”.
Using Spotify also led to an increase of music activity, with only two exceptions were there was a decrease in activity. It increased the amount of music the interviewees listened to, as well as the overall daily time music was listened to. The majority of the interviewees also expressed that they listened to more diverse music due to Spotify. Besides increase in activity and being a facilitator for new music, according to Lovisa Spotify plays another significant role that was not observed among any other interviewees:
Page | 35
“It does play a sort of leading or steering role in the music I choose to try. I think it also plays a role to actually deeper my knowledge about an artist. Because I go for the back to look at the older albums and then read their biography and maybe from there start reading more on the artist. I think it does play an educational role”.
None of the other interviewees addressed a similar topic. It was also not anticipated that Spotify would be used, or viewed, in such a way. But given the fact that there is only one interviewee using Spotify in such fashion, the sample size is too small to make any meaningful conclusions regarding this topic.
4.5 Playlist and music sharing Håkansson, et al. (2007) described the social nature of music itself, in similar fashion the interviewees experienced this with music in general and with their playlists as well. They talk about music with friends and family, shared their interests if the occasion allows it and have friends with similar music taste. Besides the social aspect there was also a strong focus on private listening, or to use Graves-Brown (2009) concept of “solitary listening”. Especially female interviewees valued their privacy strongly and were reluctant to share certain playlists openly with others. Nora stated regarding this subject that: “I don't really like to share my music. So I usually just keep it to myself. So I rarely listen to music on speakers and stuff in the apartment. It's mostly just in my own bubble”.
This is strong contrast with Joep who stated that he doesn’t care if people see what music he listens to. Male interviewees were also more likely to create playlists for and with others. Edgar has made several playlists on request from others. The reason for this is that people around him are aware of his passion for music and consider him a good source for recommendations. Alex has a collaborative playlist with his friends from Sweden that they update regularly. A collaborative playlist brings different dynamics into play as Alex and his friends have to consider each other’s taste as well. “With collaborative playlists you have to think about what other people want. And if you don't even discuss, like we've never even done, maybe next time I see these guys I will tell them: "Can we delete some songs". But since you don't discuss it, you have to pay attention... Yeah, to pay attention to other people's wishes, then it becomes a bit much”.
Page | 36
The playlist itself is rarely discussed among Alex and his friends, but there is a constant group pressure hovering over the playlist. Deleting songs is considered a social faux pas without permission of the group. Music sharing via Spotify, or in general, is primarily done out of shared interest or getting acquainted with someone else’s music taste. Both Ilona and Gerard used shared playlists to get to know another person better. Spotify had a role that it allowed social interaction without directly talking with each other. For both of them listening to the playlists gave them an insight into the musical habits of the other person of interest. A useful way to get to know a person better. There was no initial expectation that this would be mentioned by the interviewees. But even on a relative small scale that is this research, it occurred twice. Although the interviewees on several occasions relied on others for music recommendations, there were no significant indications that supports the findings by Yaniv, et al. (2010) when it comes to looking for a recommendation. None of the interviewees had a specific preference for recommendations from individuals of the same gender. The major requirement for a successful recommendation was whether or not the recommender shared a similar music taste or was aware of the music taste of the interviewee. Furthermore, there was no significant difference between male and female interviewees regarding how much they rely on music recommendations. Spotify opened up new possibilities for sharing and recommending music to others. The actual recommendation process was usually done face to face among the interviewees. The role of the parents a relative larger role in this process. Nora, Maaike, Rosa and Laura all had valued their parent’s music taste and expressed to be influenced by this as well. This could range from having a similar taste, in the case of Laura, or listening to similar music as their parents did. Rosa created a playlists specifically with her father’s favorite artist: “That [Lou Reed] is one of his favorite artists. So I heard that a lot when I was younger. Recently I heard a song from him on the radio and then I thought “oh yeah, that’s what I listened when I was young”. So then I started listening to a couple of his songs again”.
Male interviewees expressed much less connection to the music their parents listened too, with Jordh being the only one stating that he shares musical interests with his father. Overall male interviewees relied more strongly on music recommendations from friends with a similar taste in music. Throughout all discussed topics there were significant differences among the male and female interviewees. Female interviewees expressed much more emotional connection to music while the male interviewees stayed more superficial and talked usually about the kind of music they like, while avoiding talking about topics such as emotions. Male interviewees seemed to diminish the role Spotify had on their music Page | 37
consumption and what kind of role it played in their own identity formation, in comparison to the female interviewees. In the initial expectations it was not accounted for that male interviewees were more likely to express difficulties regarding expressing emotional connections to music. This observation opens up a different debate regarding the likelihood of emotional expressions among different genders. I will briefly state that these results should not have been unexpected, considering differences between emotional expression among men and women have been the topic of research, especially in the field of psychology. The following conclusion by Barrett, et al. (1998) seems to fit well with what has been observed during the interviews: “Even though they may not differ in their momentary experiences, women may attend more to their emotions when they occur, think more about them, share their emotions more with other people, and find more meaning in their emotional reactions than men. These experiences may lead women to develop a more elaborated view of themselves as an emotional being than typically occurs for men” (p. 573). While this doesn’t mean that music is inherently experienced in a different way, women may be more willing to express themselves emotionally when it comes to their music preferences.
Page | 38
CHAPTER 5 – CONCLUSION Conducting twelve interviews with Spotify users has brought up a diverse range of new insights on how Spotify is experienced on an individual level. This ranged from being merely a helpful tool in the overall music experience to a lifestyle accessory. With one interviewee expressing his loss of interest for Spotify, it demonstrates the difficulty streaming services experience when it comes to satisfying the needs for each individual streamer. The twelve interviewees brought twelve different musical backgrounds, music preferences, idea’s and visions about music. Overall Spotify is viewed positively among the interviewees. It has been integrated to such an extent that it is difficult to imagine a life without the ability to access music on such a large scale at any given time. This is all achieved in merely six years since Spotify has been active in the Netherlands. Given such a rapid transformation of the musical landscape it is not surprising that research on how these streaming services influence music consumption and experience is lacking at the moment. The attempt for this research was to give an insight on how streamers use Spotify but more importantly, their playlists. Because music is both a social and personal tool, streamers can use their Spotify accounts and playlists connected to those accounts as a tool of personal expression. Conducting interviews has given insight into the selection and creation processes of playlists and what their playlists mean to them on a personal level. The most striking results were the differences in the way Spotify is used and experienced between the male and female interviewees. Age played a smaller role, but due to the fact that there was only one user aged above 40 little significant conclusions can be drawn. The age group between 20 and 25 however, were not more active than the interviewees of a higher age group.
5.1 Findings 5.1.1 Complimentary to identity This study had to rely, for the largest part, on existing research that has been done prior to the existence of streaming services. This made it difficult to provide extensive expectations. For the user above 40 Spotify functioned as a helpful tool to an already existing and shaped musical identity, while the interviewees below this age group were so accustomed to having Spotify that it became part of their daily life in similar fashion to a smartphone. A commodity that has become difficult to live without. Unfortunately it proved to be difficult to find users of the above 40 age group that use Spotify in similar fashion as Gerard did. This age group experienced music in a different fashion (i.e. listening via physical media such as CD’s or vinyl) for the majority of their lives. For them Spotify could be a great tool that opens up a lot of new ways to find and listen to music, but it is not unreplaceable. The lower age group Page | 39
(bellow 40) differed strongly and in this group Spotify formed the predominant way they consume and experience music. However how Spotify’s role regarding identity formation works, proved to be difficult to observe in the interviews. Instead Spotify compliments an existing identity, usually referred as ‘personality’ by the interviewees. As in the case of Alex, Spotify fitted in his believe of a minimalist lifestyle where ownership is pushed more to the background in favor of borrowing, or in the case of Spotify ‘using’ music. For the remaining interviewees it was predominately a helpful tool but did not offer any significant complementary benefits to their daily life. Spotify did help the interviewees find new music and for a minority of the interviewees it helped to develop a narrowed down music taste, albeit in a more practical sense. The creation of playlists and programmed recommendations provided by Spotify allowed easy access to the favorite music of the interviewees. Tools such as Discover Weekly and ‘similar artists’ were appreciated by the majority of the interviewees and brought them into contact with new or unknown music. The majority of the interviewees relied relative strongly on automatic generated recommendations, Discover Weekly being the most prominent. While acknowledged by the interviewees that Spotify made sharing music easier, via for instance playlists and linking it with other social media, any significant increase in sharing music was not observed during this research.
5.1.2 Activity, emotion and gender Spotify did clearly increase music activity for six interviewees, for the rest it remained the same or it had only a slight influence on activity. Spotify was consumed everyday by all interviewees, with Joep being the only exception. The overall amount of different types of music consumed increased as well for the interviewees, which was predominately due to the accessibility of the platform according to the interviewees. It did not necessarily change any music preferences the interviewees had before the use of Spotify. Although the interviewees listened to more diverse music, the interviewees on average still expressed favorite bands or artists that were listened to most frequently. The music other than that of their favorite artists is more fleeting and skipped through regularly. A couple of songs from one artists and then quickly moving on to the next. Anything that is not considered worthy enough will be listened to once or twice, but will not end up in one of the playlists the interviewees created. The Spotify playlist has become more than a representation of musical taste. They are used depending on mood, connected to personal memories or seen as a representation of one’s personality. The analysis demonstrated that these feelings most strongly occur among women. This research has shown a significant difference in the way Spotify is experienced among different genders. That difference is not only prevalent looking at the
Page | 40
playlists, but also on the social aspect of music. Women were more likely to use Spotify for solitary listening as their music preferences were considered a private affair. While sharing did occur among both men and women, men proved to be more indifferent and cared little if others could see their playlists or what they were listening to. While difficult to generalize, previous research has demonstrated that women are more likely to use music as “emotional management”, that is to say listening to music as a mood enhancer or “tranquilizer” (Wells & Hakanen, 1991). With that respect, this research has brought up similar results as the female interviewees created mood enhancing (workout, study) or tranquilizing (relaxation, sleep) playlists.
5.2 Research question and discussion How does the playlist creation process take place and what meanings do streamers associate with the music on their playlists? Rather than providing an answer to this question, this research provides a start for research regarding similar topics. The startup phase looks to be finished and Spotify is here to stay (or at the very least streaming music in general). It is likely that Spotify will receive more academic research within the coming years. Provided with twelve different opinions and views about music, Spotify and playlists, several conclusions can be drawn. Male interviewees attached less meaning to the playlists as they predominately remained collections of their music preferences or created out of function. That is to say for a vacation or specific event. For the male interviewees playlists were not a representation of their personality in any specific way. Contrasting the female interviewees who predominately created playlists, on top of music preferences, based on mood and/or activity. The female interviewees were more able to express that their playlist collection is somewhat a representation of their personality, while this was not necessarily the case for all the female interviewees. Because the interviewees were no longer limited by physical media such as CD’s there was less incentive to listen to entire albums. Therefore, the playlists rarely focused on a single artist or band. Rather than that, playlists covered entire genres. There was no significant similarities observed in the playlists among the interviewees, besides the overall differences among male and female interviewees. Each playlists collection was unique and each interviewee applied their own criteria on what was worthy enough music to be added to a playlist. A playlist can start as something functional related (i.e. for a party or activity), where the creator simply associates certain music with an activity. This is based on personal criteria and were not interchangeable among the interviewees. While the interviewees had consensus over the fact that a ‘workout’ playlist needed up tempo music, this does not mean that there were many similarities regarding music selection. Mood or memory related playlists, were more likely to be created out of an organic process. Not purposely creating a Page | 41
playlist, but already associating music prior to the playlist to a certain memory or feeling. That slowly formed into a playlists, as over time songs were added or removed if necessary. The playlists created by the users are not necessarily a collection in similar fashion as one can collect CD’s for instance. There are not put on display and are more easily discarded as there are digital representations of taste, not physical. However, user created Spotify playlists are unique in the sense that the reasons for creating them can be as diverse as an individual’s music taste. Mood, memory, activity, favorites, all these topics played a role in the creation process. Beneath the surface of something that can be so easily created, is the musical identity of the user. In that sense the playlist has value to the user as it can be a tool for managing emotions, helping to concentrate or reliving memories. Considering this the ability to create playlists out of an extensive library of music allows users to think about music in a different fashion. While this was certainly possible prior to Spotify, the accessibility of the platform allows users to think more consciously about what music they like, what it means to them and how it can benefit them.
5.3 Limitations The limited sample size of twelve interviewees cannot account for generalizable conclusions on Spotify behavior on larger levels, such as region or country. On top of that there has been a focus on people from the Rotterdam area only. The research did include three nationalities (Dutch, Swedish and Norwegians) but were not accounted for as variables. However, there were no significant differences observed between the Dutch and Scandinavian interviewees in the way Spotify was used. Therefore this was not elaborated further as the non-Dutch interviewees were all living and working/studying in Rotterdam. The second limitation regards the interview themselves. Prior to the interviews an interview guide was created with the necessary questions to answer the main research question. However questions related to the role of Spotify or the meaning behind playlists proved to be difficult to answer by some interviewees. This could be due to the fact that these questions appeared ‘out of nowhere’ for some interviewees. The conversations did not necessarily led up to these questions. Nor is it something the interviewees actively think about when they look at their playlists. Perhaps this could have been avoided by instructing the interviewees more clearly regarding the expectation of the interview. By instructing them, prior to the interview, to go through their playlists and think about why they have created their playlists, they could potentially have a better overview of their own reasoning.
Page | 42
5.4 Future research Potential further research could focus more on the influences Spotify has on listening behavior of the individual consumer. As Spotify (and its competition) continues to rapidly increase, it looks like it is quickly on its way to become the definitive way to consume music. As each streamer uses Spotify in his or her personal way, what it does to the amount of music we consume whether that be in hours used or how many different artists listened to, is severely lacking in academic research. This research provides a starting point insofar that it demonstrates that playlists are a well thought organic process. They are created for specific purposes and edited if the situation demands it. Providing the fact that there were significant differences observed between male a female interviewees, this gender difference is something that can also be taken into account in future research. The question there could be if males truly experience less emotional attachment to playlists, or even music in general. While this was not an initial expectation, the stark differences between how music is emotionally experienced among different genders was even observed in a relative small sample size focused on Spotify playlists.
Page | 43
REFERENCES Adermon, A., & Liang, C. Y. (2014). Piracy and music sales: The effects of an anti-piracy law. Journal of Economic Behavior & Organization, 105, 90-106. Aguiar, L., & Waldfogel, J. (2015). Streaming Reaches Flood Stage: Does Spotify Stimulate or Depress Music Sales (JRC96951)? Retrieved from National Bureau of Economic Research: http://www.nber.org/papers/w21653 Alonso, M. B., & Keyson, D. V. (2005, April). Music Cube: Making Digital Music Tangible. In CHI'05 Extended Abstracts on Human Factors in Computing Systems, 1176-1179. doi: 10.1145/1056808.1056870 Artist unkown. Spotify logo, black background [digital image]. Retrieved from: https://press.spotify.com/us/images/ Artist unkown. Spotify logo, transparant [digital image]. Retrieved from: https://press.spotify.com/us/images/ Avdeeff, M. (2012). Technological engagement and musical eclecticism: An examination of contemporary listening practices. Participations: Journal of Audience and Reception Studies, 9(2), 265-285. Bahanovich, D., & Collopy, D. (2009). Music experience and behaviour in young people (UK Music) (pp. 1-23). Retrieved from University of Hertfordshire: http://i.contentreview.com/s/ebf223550adb6d7c64fa176200d66ec8.pdf Barrett, F. S., Grimm, K. J., Robins, R. W., Wildschut, T., Sedikides, C., & Janata, P. (2010). Music-evoked nostalgia: affect, memory, and personality. Emotion, 10(3), 390-403. Barrett, L. F., Robin, L., Pietromonaco, P. R., & Eyssell, K. M. (1998). Are women the “more emotional” sex? Evidence from emotional experiences in social context. Cognition & Emotion, 12(4), 555-578. Becker, H. S. (1996). The epistemology of qualitative research. In Jessor, R., Colby A. & Shweder, R.A., Ethnography and human development: Context and meaning in social inquiry (pp. 53-71). Bryson, B. (1996). “Anything but heavy metal": Symbolic exclusion and musical dislikes. American sociological review, 61(5), 884-899. Chamorro‐Premuzic, T., & Furnham, A. (2007). Personality and music: can traits explain how people use music in everyday life? British Journal of Psychology, 98(2), 175185. Cunningham, S. J., Bainbridge, D., & Falconer, A. (2006). “More of an art than a science" Supporting the creation of playlists and mixes. Seventh International Conference on Music Information Retrieval (ISMIR) (pp. 240- 245). Victoria, Canada. DeNora, T. (1999). Music as a technology of the self. Poetics, 27(1), 31-56. Page | 44
Frith, S. Music and identity. In Hall, S. & Du Gray, P., Questions of cultural identity (pp. 108127). Gibson, C., Connell, J. (2002) Sound tracks: Popular music identity and place. New York: Routledge: Taylor & Francis Group. Gibson, C., Connell, J. (2005). Music and tourism: On the road again. Clevedon: Channel View Publications Goldmann, M., & Kreitz, G. (2011, August). Measurements on the Spotify peer-assisted music on demand streaming system. In Peer-to-Peer Computing (P2P), 2011 IEEE International Conference on (pp. 206-211). doi: 10.1109/P2P.2011.6038737 Graves-Brown, P. (2009). Nowhere man: urban life and the virtualization of popular music. Popular Music History, 4(2). Hagen, A.N. (2015). The Playlist Experience: Personal Playlists in Music Streaming Services. Popular Music and Society 38(5), 1-21. doi: 10.1080/03007766.2015.1021174 Hanslick, E., & Payzant, G. (1986). On the musically beautiful: A contribution towards the revision of the aesthetics of music. Indianapolis: Hackett Publishing. Håkansson, M., Rost, M., & Holmquist, L. E. (2007). Gifts from friends and strangers: A study of mobile music sharing. In ECSCW 2007 (pp. 311-330). London, Springer. Hargreaves, D. J., & North, A. C. (1999). The functions of music in everyday life: Redefining the social in music psychology. Psychology of music, 27(1), 71-83. Firth, S. (1996). Music and Identity. In Hall, S. & Du Gay, P. Questions of Cultural Identity (p. 108-150). London: Sage. Hays, T., & Minichiello, V. (2005). The meaning of music in the lives of older people: A qualitative study. Psychology of music, 33(4), 437-451. Hetland, L. (2000). Listening to Music Enhances Spatial-Temporal Reasoning: Evidence for the "Mozart Effect". Journal of Aeshtetic Education, 34(3/4), 105-148. Jupp, V. (2006). The Sage dictionary of social research methods. London: Sage. Kotsopoulou, A., & Hallam, S. (2010). The perceived impact of playing music while studying: age and cultural differences. Educational Studies, 36(4), 431-440. Legard, R., Keegan, J., & Ward, K. (2003). In-depth interviews. In J. Ritchie & J. Lewis (Eds.), Qualitative research practice: A guide for social science students and researchers (pp. 138-169). Lonsdale, A. J., & North, A. C. (2011). Why do we listen to music? A uses and gratifications analysis. British Journal of Psychology, 102(1), 108-134. Payne, E. (1980). Towards an understanding of music appreciation. Psychology of Music, 8(2), 31-41. Peretz, I. (2006). The nature of music from a biological perspective. Cognition, 100(1), 1-32. Page | 45
Premkumar, G. P. (2003). Alternate distribution strategies for digital music. Communications of the ACM, 46(9), 89-95. Rentfrow, P. J. (2012). The role of music in everyday life: Current directions in the social psychology of music. Social and personality psychology compass, 6(5), 402-416. Ritala, S. (2013). Pandora & Spotify: Legal Issues and Licensing Requirements for Interactive and Non Interactive Internet Radio Broadcasters. IDEA, 54(1), 23-53. Robinson, J. (2005). Deeper than reason: Emotion and its role in literature, music and art. New York: Oxford University Press Ruud, E. (1997). Music and identity. Nordic Journal of Music Therapy, 6(1), 3-13. Simon, R. W., & Nath, L. E. (2004). Gender and Emotion in the United States: Do Men and Women Differ in Self‐Reports of Feelings and Expressive Behavior? American journal of sociology, 109(5), 1137-1176. Song, Y., Dixon, S., & Pearce, M. (2012, June). A survey of music recommendation systems and future perspectives. 9th International Symposium on Computer Music Modeling and Retrieval. Queen Mary University of London, London. Retrieved from: http://cmmr2012.eecs.qmul.ac.uk/sites/cmmr2012.eecs.qmul.ac.uk/files/pdf/papers/c mmr2012_submission_42.pdf Sloboda, J. A., O'Neill, S. A., & Ivaldi, A. (2001). Functions of music in everyday life: An exploratory study using the Experience Sampling Method. Musicae Scientiae, 5(1), 932. Wells, A., & Hakanen, E. A. (1991). The emotional use of popular music by adolescents. Journalism & Mass Communication Quarterly, 68(3), 445-454. Yaniv, I., Choshen-Hillel, S., & Milyavsky, M. (2011). Receiving advice on matters of taste: Similarity, majority influence, and taste discrimination. Organizational Behavior and Human Decision Processes, 115(1), 111-120. Zhang, B., Kreitz, G., Isaksson, M., Ubillos, J., Urdaneta, G., Pouwelse, J., & Epema, D. (2013). Understanding user behavior in Spotify. INFOCOM, 2013 Proceedings IEEE, 220-224. doi: 10.1109/INFCOM.2013.6566767
Page | 46
APPENDIX 1 – INTERVIEW GUIDE Before the start of the interviews I am familiar with the following background information: gender, nationality and being an active Spotify user who has created his or her own playlist(s). Therefore there are no question such as: Do you create your own Spotify playlists? All the interviewees are active on Spotify on a daily basis, have created one or more playlists and/or listen to automatic generated playlists by Spotify (such as ‘summer’ or ‘Christmas’ playlists). I will avoid describing automatic generated playlists by Spotify as such, for reasons that the words “automatic generated” might have a negative connotation to it. Therefore I will just call them “Spotify playlists” within the interviews. A Spotify user is likely to be familiar with these automatic generated playlists and therefore should not have a problem understanding the questions. Furthermore age and educational level will also be asked for either prior or at the beginning of the interview. The interviews will be conducted in Dutch and English if the interviewee is not able to conduct an interview in Dutch. The questions cover a wide variety of topics that range from music preferences, listening behavior prior to Spotify, recommendations and the sharing of music with others. These questions will form the overall structure of the interview. Depending on the interviewee some questions might differ or be altered as the interview progresses.
APPENDIX 1.1 Interview introduction First and foremost, I want to thank you for agreeing with this interview. This interview will be part of my master thesis and I am going to talk with you about your music taste and your Spotify behavior. The results from the interview will be used to answer my research question. I will talk with you about the playlists you have created, how those playlist came to be and what they mean to you. The interview will be recorded, audio only. The information discussed during the interview will be handled carefully and the audio file is for transcription purposes only. I will start off with some basic questions to discover your music taste. There is no right or wrong answer and you can take your time to think about the questions and your answers. If possible I would also like to see some of your playlists you have created on Spotify, so we can browse through them. We can talk about some of the music you added to the playlist and why you added it. But first I would like to start a with your music preferences in general.
Page | 47
APPENDIX 1.2 Interview questions TOPIC: Music in daily life and listening behavior 1. What music do you like? 2. When do you listen to music? 3. How do you listen to music? 4. How often do you listen to music? 5. What role does music play in your daily life? 6. How important is music to you? Why? 7. Do you visit concert/festivals or any other music related events? 8. Do you make music yourself?
TOPIC: Streaming and playlists 9. How did you listen to music before Spotify? 10. When did you start streaming music on Spotify? 11. How did you get introduced to streaming? 12. How many minutes/hours a day do you use Spotify? 13. Did your music activity increase with the rise of streaming services? 14. What are the advantages of streaming music? And the disadvantages? 15. When do you create a playlist? 16. How do you select the music you want to add to your playlist? 17. What do the playlists you have created mean to you? 18. On what occasions do you listen to a playlist? 19. Do you create playlists to fit a particular occasion?
TOPIC: Recommendations and sharing 20. How do you discover or look for new songs to listen to on Spotify? 21. What do you think of recommendations that your friends or family give you? 22. And what about Spotify recommendations? 23. How do you share your playlists and music in general with others? 24. Have you created a playlist with someone else? 25. When do you share your music with others?
Page | 48
TOPIC: Role of Spotify 26. What role does Spotify play in the way you experience music? 27. To what extent has Spotify changed the way you experienced music previously? 28. What would you like to change to Spotify?
TOPIC: Meaning behind playlists 29. When you talk with others about your music/playlist collection, what do you discuss? 30. What does your music/playlist collection says about you? 31. To what extent do your music preferences connect with your lifestyle?
Page | 49
APPENDIX 2 – Transcriptions Interview transcriptions available on request.
Page | 50