1. Introduction
This study will examine the impact of Spotify on music consumers and music discovery in the digital age. What used to be an industry-led by physical interactions and profitable copies of musical material, the economic climate of the music world has now become heavily saturated and easily digestible through streaming services like Spotify, Apple Music, and YouTube (Parys, Wares, 2022). Officially launching in 2006, subscription-based streaming service Spotify had accumulated 188 million subscribers as of 2022 and averaged 60,000 songs/podcasts uploaded daily, acquiring a vast catalog (Samanta, 2023). Artists can quickly release music through distribution companies such as Distrokid to make it available on numerous streaming platforms. Since the massive popularity of Spotify, music listeners have expressed concerns about the ability of algorithmic recommendations to effectively enhance musical discovery experiences (Freeman et al., 2022). Progressive artificial intelligence and new-age data software have revolutionized the music industry for all parties. Technologically speaking, streaming platforms have surged with power over their service’s consumption (Werner, 2020). Spotify’s features have cultivated a digital environment beyond undemanding music streaming but rather a mechanism for individual identity manifestation (Sesigür, 2021). The streaming platform has earned its reputation as a standard music tool over the past decade. The rapid pace and literal data computations of Spotify’s curation software raise concerns from consumers who personally identify with their music library (Freeman et al., 2022). For example, if user one allows user two to play music on user one’s account, but the music does not reflect their personality, user one could become concerned about the automated software interpreting their future recommendations and the following digital footprint (Freeman et al., 2022).
As the year ends, Spotify will release “Spotify Wrapped,” which highlights what songs, genres, and artists were listened to the most by an individual user. Typically, users will share their statistics and results with their acquaintances or on social media, which is viewed as a way of establishing personal identity and musical taste (Freeman et al., 2022). Thus, the listener can see the data processor as either a friend or a foe, which is more complicated than what appears on the surface (Webster, 2023). For relevance and popularity purposes, this study will focus specifically on Spotify, given that the company is the leading streaming platform by a considerable margin and concerns about algorithmic functionality have surfaced. This study aims to examine and analyze the awareness and attitudes of Spotify users regarding the effectiveness of the streaming platforms’ algorithms, features, and music discovery capabilities.
2. Literature Review
On a technical level, downloading music electronically originated in 2000 with the creation of Napster, a music piracy website (Mann, 2000). This site allowed over 20 million listeners to download music files onto their computers for free with no royalties given to the artists (Mann, 2000). Given Napster’s failure after numerous lawsuits regarding royalties, companies such as Apple Music and Spotify presumed a similar business construct but required subscriptions if the listener wanted an ad-free experience. Between 1999 and 2016, physical music sales decreased yearly, resulting in an all-time low and all-time high for digital music consumption (Kubala, Szymkowiak, 2020). Studies have indicated that streaming platform users fear the quality of discovery algorithms, distorting the amount of music exposure (Aguado et al., 2022). However, Spotify has evolved beyond Napster’s initial mission. Whereas Napster was a catalyst in allowing users to digitize music through .wav or mp3 files, Spotify is a vessel for discovery and personalization. The companies’ origin began making music accessible in mass quantities but progressed towards establishing several functions that entice users to increase interactions (Freeman et al., 2022; Webster, 2023).
In 2014, Spotify incorporated recommendation technology, the Echo Nest, into their operation system (Chodos, 2020). Automated software became Spotify’s foundational perk, inclining users to feel personally addressed. In one study, a user described Spotify as “I would say that it is almost like the eldest sibling that knows you better than you know yourself and therefore can recommend things that may be from left field. And sometimes you kind of hear those things, and you’re like, ‘I can see where you’re coming from'” (Freeman et al., 2022). This user’s description can be considered a transparent intimacy between humans and machines (Freeman et al., 2022). Once Spotify included features such as the curated playlists made for users Discover Weekly or Daily Mix, users no longer felt obligated to perform the same amount of listening decisions. These features are products of data based on the user’s listening habits and patterns, thus creating functions that shape the listener’s decisions. When Spotify launched this function in 2015, the press release depicted, “It is like having your best friend make you a personalized mixtape every single week” (Hamilton, 2019; Spotify, 2015). The rhetorical devices used in this statement likely induced the sense of intimacy previously mentioned by a Spotify user. However, I propose that the explicit automated playlist advertisement rather than the extensive music collections available through the “browse” option has the potential to restrain users from expanding beyond the measures of Spotify’s instantly gratifying features, thus creating a cyclical dependency on the application’s algorithm (Freeman et al., 2022, p. 3). Suppose the algorithm controls the material on playlists and radios based on user data. In that case, the software is responsible for the amount of music discovery occurring in a scenario where the user exclusively listens to the automated functions. The demonstrated ideology is not to say Spotify’s playlists do not assist users in discovering new music. Spotify’s playlists shape the music’s presentation to consumers (Werner, 2020).
It would be presumptuous to claim the format in which Spotify presents music to its users as a negative quality. The only limitations when using Spotify depend on whether artists have published their music on the platform. For instance, fans of Neil Young and Joni Mitchell cannot stream the two artists’ original music due to removing the entirety of their original material in protest of accusing Spotify of spreading false information about COVID-19. However, Spotify still offers features that allow users to seek out artists. Functions such as related artists, browse, artist radio, and playlists made by curators worldwide are non-algorithmic discovery functions (Werner, 2020). The related artist function displays artists that are either similar musically or share similarities socially. The browse function leaves the user with an oasis of opportunities and a broad, open-ended catalog. Artist radios act similar to the related artist function but seemingly play related artists without the user’s efforts (Werner, 2020). Therefore, Spotify reduces the music discovery limitations outside of algorithmic functions.
Once again, the objective of this study is not to discuss Spotify’s limitations, as it would be invalid. The objective is to analyze the attitudes of Spotify users towards algorithms and if the users deem the streaming platform satisfactory. To preface the course of this research, Freeman et al. (2022) recorded feedback from Spotify users that is useful moving forward. “I have been on a home screen previously and thought to myself, ‘Oh, this is a reasonable interpretation from someone who hasn’t met or doesn’t know me as a person’. And then if I would listen to other genres, like I also really like rap, but nowhere near to the same level as those other three [main] categories. And then it felt like Spotify was like, ‘You’d been listening to rap again, haven’t you?’. And that would be like Daily Mix 4…” Ryo (Freeman et al., 2022).
The following quote displays the data software recording the user’s behavior and reacting immediately through an algorithmic Daily Mix playlist. The only way the user can communicate their pleasure with that result would be to react accordingly, such as listening to more rap if the user prefers to receive more rap on the playlists, and vice versa. However, the user felt the platform “betrayed” his taste in music and struggled to trust the algorithm to produce playlists that accurately depict his taste (Freeman et al., 2022, p. 7).
“So if I press ‘like’ on something, I think, “this is good, you’re learning and I want you to take note of this”. So if I’m at work it recommends music that I like, and I’m distracted, I’ll go back and look through the thing from half an hour ago and go and like those songs because I want it to do the job.” Ryo (Freeman et al., 2022).
Though this quote refers to passive listening, it still highlights that the user must train the algorithm to adapt to their behavioral patterns. By following this act, the user exerts more effort into a function than desired, hoping it will be algorithmically rewarding (Freeman et al., 2022). To adjust the patterns of the algorithm, the user must be aware of the algorithm’s reactionary behaviors in correlation to the user’s behaviors. Analyzing the data-obtaining process and its structure would be unnecessary, given that it would vary per user; therefore, it is valid to record algorithmic patterns to predict reactions (Werner, 2020). According to the following studies, framing theory suggests that users’ attitudes towards algorithmic behavior depend on their sole purpose of using the Spotify app.
3. Hypothesis and Theory
Perhaps roughly less than a decade ago, the algorithmic studies as a function would be more crucial to discuss as technology intimidatingly progressed (Datta et al., 2017). Modern media has advanced beyond the demand to understand technology. The primary conflict is now understanding how society is supposed to react, and the focus has shifted from technological impact to socio-technical comprehension.
However, according to the literature studies, Spotify users tend to be aware of algorithmic behavior to the extent that the users attempt to train the data processor to accurately display the user’s music taste. Objectively inaccurate depictions of users’ music tastes and reactions to user behavior lead to users actively attempting to discover music without algorithmic assistance. Framing theory suggests the user’s reaction in the second example of the literature review depends on the streaming software and is an attempt to communicate their desired correction. This example indicates a negative attitude towards algorithmic functions. In contrast to the user’s reaction, the same user admitted the algorithm accurately displayed their music taste through the application’s home screen selection. This reaction indicates a positive attitude, given that the algorithm accurately displayed the user’s music taste and provided appropriate options to assist them. Therefore, the users’ attitudes towards Spotify’s algorithms depend on their primary purpose for using the app.
H: Spotify users with a positive attitude towards algorithmic behavior prioritize passive music discovery and utilize algorithmic patterns. In contrast, those with a negative attitude towards algorithmic behavior prioritize active music discovery and unlimited music access.
4. Variables
This study aims to decipher Spotify users’ attitudes in parallel to their purpose for using the platform to imply that the dependent variable would be the audiences’ attitudes towards algorithmic patterns. The independent variable would be the audiences’ reason for using Spotify as their music listening platform. The unit of analysis in this study will be Spotify users. Though all users may not exclusively use Spotify for unlimited music access or music discovery, the methods stated in Section 5 will allow participants in the study to state the other reasons for usage to measure the rate at which users relate to the hypothesis. However, users who represent applicable data to the study will be prioritized.
5. Methods
In order to further understand the attitudes of music consumers in the digital age, an extensive survey will be conducted, the results of which will indicate the attitudes of streaming platform audiences of different demographics. The unit of analysis for this research will be the previously mentioned audiences, as they will be asked questions regarding the effectiveness or ineffectiveness of their music listening and discovery methods. This questionnaire will consist of 50 randomly selected participants filling out an anonymous questionnaire of closed-ended questions with mutually exclusive categories and an option that reads “Other” to produce uniform results. I will conduct this survey as a purposive sampling method. I will create a Google Form document where the data will be transferred into an Excel sheet in order to analyze the data correctly. Questions will only pertain to the individuals’ perceptions of music streaming platforms or their preferred music consumption methods to prevent convolution in the results, as well as their age and gender. The objective of the survey is to further understand, examine, and analyze the attitudes of average music listeners and decipher the effectiveness of streaming platforms and their ability to enhance the music discovery experience in the digital age.
6. Results
Section 6.1
Upon receiving 50 responses to the 15-question survey on Google Forms, I concluded the questionnaire to finalize the results. This survey was conducted anonymously and transferred between individuals I shared no prior communication to increase diversity. Questions 1-8 consisted of multiple-choice answers with an option that reads “Other.” Questions 9-12 allowed respondents to select their extent of satisfaction and awareness. By deviating the formatting of the questions, questions 13-14 allowed respondents to answer on a scale of 1-10. Conclusively, question 15 allowed an open-ended response where respondents would describe their attitudes towards Spotify algorithms and functions in their own words.
As mentioned, I proposed that Spotify users with a positive attitude toward algorithmic behavior prioritize passive music discovery. In contrast, those with a negative attitude towards algorithmic behavior prioritize active music discovery. In order to properly test this hypothesis, results that apply to the respondents’ purpose for using Spotify will be those who prioritize unlimited music access and music discovery.
In prefacing the results, 60 percent of respondents identified as male, while the remaining 40 percent identified as female. Sixty-eight percent of respondents were between the ages of 23 and 29, and 24 percent were between the ages of 16 and 22, leaving the remaining 8 percent of respondents to be 30 years old or more (Figure 6.1).
Figure 6.1. Pie chart indicating the age range of respondents.
Given the importance of how users are affected by Spotify consumption daily, the respondents were asked an approximation of how often they use Spotify within a standard 7-day week. According to the results, 74 percent of respondents use Spotify daily, 24 percent 4-6 days, and the remaining 4 percent use Spotify 2-3 days a week (Figure 6.2).
Figure 6.2. Pie chart indicating the frequency in which consumers use Spotify.
After establishing the usage frequency of the respondents, the survey proceeded to examine the habits and purposes of using Spotify. Ninety-six percent of users pay for the premium ad-free monthly subscription, while 4 percent engage with the ad-included free subscription. Understanding the respondents’ purpose for using Spotify contributes to the original objective of examining their attitudes. In contrast to past comparisons of Napster exclusively being an outlet for unlimited music downloading, as opposed to the multi-faceted Spotify catalyzing music discovery, 74 percent of respondents claim to use Spotify due to its unlimited music access. Sixteen percent of respondents prioritize music discovery, and 4 percent utilize all purposes (Figure 6.3). Most respondents utilize the platform for its musical catalog, and 50 percent of respondents typically generate the most streams from their created playlists. The second-highest method of generating streams resulted in Spotify’s algorithmic playlist.
Figure 6.3. Pie chart indicating respondents’ purpose for using Spotify.
The ratio between the audiences’ purpose for using Spotify resulted in a wide margin. However, the attitudes of both categorical respondents varied. According to the hypothesis, users with a positive attitude towards algorithmic patterns prioritize passive music discovery and utilize the algorithm. In contrast, those with a negative attitude towards algorithmic patterns prioritize active music discovery and unlimited music access. Figure 6.4 represents the overall results of respondents’ methods of discovering music.
Figure 6.4. This pie chart refers to the methods each respondent uses for music discovery.
According to the results, 74 percent of respondents utilize algorithmic functions, such as algorithmic playlists, DJ functions, related artists’ functions, artists’ radios, or smart shuffle. The remaining 26 percent utilizes functions that require active searching.
In order to further understand the attitudes of those who pertain to the hypothesis, I omitted 5 participants who did not claim to use Spotify for unlimited music access or music discovery.
Section 6.2 Participants who prioritize music discovery
Regarding 16 percent of participants who use Spotify for music discovery, 25 percent claim most of their streams originate from algorithmic functions such as algorithmic playlists or the related artist’s function. The remaining 75 percent of streams originate from non-algorithmic functions such as Spotify’s editorial playlist, their libraries, or their playlists (Table 6-1).
| Main source of streams | Ratio |
| Spotify Algorithmic playlists | 12.50% |
| Users’ personal playlists | 50% |
| Users’ personal libraries | 12.50% |
| Related Artists Function | 12.50% |
| Spotify Editorial Playlists | 12.50% |
Table 6-1. This table indicates the main source of streams from users who prioritize music discovery.
Table 6-1 shows that less than a third of users in the 16 percentile commonly practice algorithmic functions regularly. However, these results could indicate that their usage proceeds with their usual discovery methods, represented in Table 6-2.
| Discovery Methods | Ratio |
|---|---|
| Spotify Algorithmic playlists | 62.50% |
| Related Artists function | 25% |
| Artist radios | 12.50% |
Table 6-2. This table represents the discovery methods used by those who prioritize music discovery.
Spotify algorithmic playlists cater to the user’s listening patterns to simplify decision-making. Playlists typically include artists commonly listened to, with the occasional artist that would likely be found on an artist’s radio. Among the 16 percent of respondents who prioritize music discovery, 37.5 percent consider themselves extremely aware, 50 percent moderately aware, and 12.5 percent indifferent. Given their awareness and the results of their listening behaviors, the survey then examined their satisfaction with algorithmic patterns.
| Spotify Music Discovery Experience | Ratio |
| Extremely satisfactory | 50% |
| Moderately satisfactory | 37.50% |
| Indifferent | 12.50% |
| Moderately unsatisfactory | 0% |
| Extremely unsatisfactory | 0% |
Table 6-3. This table indicates the satisfaction in regards to Spotify’s music discovery experience amongst those who prioritize it.
According to Table 6-3, respondents amongst the 16 percent that prioritize music discovery are generally satisfied with their experience, except the 12.5 percent who are indifferent. Participant #50 displayed positive responses, was moderately satisfied with algorithmic patterns and overall experience, and frequented the related artist function.
“I think that Spotify is a good way to find new artists. I’m sure there are other ways, but the artists they have recommended for me have all been great and led me down rabbit holes of new music.”
~ Participant #50
Though most participants are pleased, various open-ended responses have indicated slight critique. Participant #47 claims to be highly aware of algorithmic patterns, moderately satisfied with Spotify’s music discovery experience, and frequents algorithmic playlists.
“I think they do a really good job at recommending songs based on my listening history. They make it really personal and it’s a much better experience than any other music platform I’ve used and I’ve discovered the most music on spotify. However, some of the recommendations can be very repetitive. I wish there was a bit more variety with the song and artist recommendations.”
~ Participant #47.
| Attitudes Towards Algorithm Adequacy | Ratio |
| Extremely satisfactory | 12.50% |
| Moderately satisfactory | 50% |
| Indifferent | 25% |
| Moderately unsatisfactory | 0% |
| Extremely unsatisfactory | 12.50% |
Table 6-4 indicates the respondents attitudes towards the adequacy of the algorithmic patterns.
In Table 6-4, 62.5 percent are satisfied with how adequately the algorithm reacts to their behaviors. Once the respondents’ attitudes towards the adequacy of the algorithm were recorded, the survey required them to scale from 1-10 on how well Spotify assists them in discovering new music, resulting in a final average of 7 out of 10. On a scale of 1-10, rating the extent to which Spotify restrains the respondents from discovering the music, the results were an average of 2 out of 10. Therefore, according to the results, Spotify users who prioritize music discovery frequently utilize algorithmic functions and display a positive attitude towards said functions.
Section 6.3 Participants who prioritize unlimited music access
Seventy-four percent of the respondents claimed to use Spotify to access unlimited music. Since this portion of the study includes significantly more participants, the results are equally imperative but show more variation. 67.6 percent of these respondents use Spotify every day of the week, 29.7 percent use it for 4-6 days, and 2.7 percent use it for 2-3 days. The streaming platform remains a prominent tool for most respondents of the unlimited music access category. Results show that 71.5 percent of these listeners accumulate most of their streams from non-algorithmic functions, whereas 26.5 percent operate through Spotify’s algorithmic playlists. However, 67.5 percent indulge in algorithmic functions to discover music, as shown in Table 6-5.
Table 6-5 indicates the methods in which users who prioritize unlimited music access discover music on Spotify.
| Discovery Methods | Ratio |
| DJ Function | 5.40% |
| Algorithmic Playlists | 32.50% |
| Other users’ playlists | 16.20% |
| Related Artists feature | 5.40% |
| Artist radio function | 13.50% |
| Social Media | 2.70% |
| Editorial Playlists | 5.40% |
| Smart shuffle function | 10.80% |
| Browse function | 8.10% |
In contrast to the original hypothesis, 74 percent of users display a higher interest in algorithmic functions than anticipated. 67.5 percent of these users employ algorithmic discovery methods. However, more participants in this category have recorded algorithmic unawareness (Table 6-6).
| Algorithm Awareness | Ratio |
| Extremely aware | 16.20% |
| Moderately aware | 56.80% |
| Indifferent | 8.10% |
| Moderately unaware | 8.10% |
| Extremely unaware | 10.80% |
Table 6-6 displays the extent of awareness amongst users who prioritize unlimited music access on Spotify.
According to Table 6-6, 73 percent of these users claim to be aware of algorithmic patterns, and this coincides with the 67.5 percent that employ algorithmic operations. Once the awareness rate was recorded, the survey began to discover attitudes and levels of satisfaction towards Spotify’s discovery experience. 64.3 percent claimed to be highly/moderately satisfied, 13.5 percent indifferent, and 16.2 percent moderately unsatisfied.
Though the majority resulted in satisfaction, Participant #31 who answered moderately unsatisfied offered critique.
“I wish that I knew how or Spotify automatically refreshed my playlists with music that is better adapted to my music tastes. The enhanced option for playlists is decent sometimes but often the songs do not fit the vibe of the playlist I made.”
~ Participant #31
Participant #40, who mostly listens to their playlists but claims to be moderately satisfied, also provided critique.
“I feel that discover weekly has had a downfall but I appreciate
the DJ function and the new daily playlist function.”
~ Participant #40
On a scale of 1-10 regarding Spotify’s effectiveness in music discovery, the answers of those who prioritize unlimited music access average 7.4 out of 10. Regarding the level of Spotify’s restraint on music discovery, these participants’ answers average 3.3 out of 10.
Figure 6.5 indicates the results of all participants and how they rate how well Spotify helps them discover more music.
Figure 6.6 indicates the overall results of the extent Spotify restrains them from discovering more music.
These results imply small-scale variation in participants’ overall attitudes and score closely to the average of both categories.
Section 7. Conclusion
This study hypothesized that Spotify users with a positive attitude towards algorithmic behavior prioritize passive music discovery and utilize algorithmic patterns. In contrast, those with a negative attitude towards algorithmic behavior prioritize active music discovery and unlimited music access. Framing theory would suggest that Spotify’s users’ purpose for usage would reflect their attitudes towards algorithms. According to the studies, there is not a substantial difference between those who prioritize unlimited music access and those who prioritize music discovery. Though the results indicate a significant margin between the two categories, the overall depiction in the study implies that most participants are satisfied, utilize algorithmic functions, display positive attitudes, and are comparatively aware of Spotify’s algorithm regardless of their purpose.
Sources
Arenal, A., Armuña, C., Ramos, S., Feijoo, C., & Aguado, J.-M. (2022). Giants with feet of clay: The sustainability of the business models in music streaming services. El Profesional de La Información. https://doi.org/10.3145/epi.2022.sep.09
Cheţan, A., & Iancu, I. (2023). The role of visual identity in music perception. KOME, 11(1), 110–128. https://doi.org/10.17646/kome.75672.94
Chodos, A. T. (2020). Solving and dissolving Musical affection: A critical study of spotify and Automated Music Recommendation in the 21st Century (dissertation). ProQuest LLC, Ann Arbor, MI.
Datta, H., Knox, G., & Bronnenberg, B. J. (2017). Changing their tune: How consumers’ adoption of online streaming affects music consumption and discovery. Marketing Science, 37(1), 5–21. https://doi.org/10.1287/mksc.2017.1051
Edlom, J. (2022). The Engagement Imperative: Experiences of Communication Practitioners’ brand work in the music industry. Media and Communication, 10(1), 66–76. https://doi.org/10.17645/mac.v10i1.4448
Freeman, S., Gibbs, M., & Nansen, B. (2022). ‘Don’t mess with my algorithm’: Exploring the relationship between listeners and automated curation and recommendation on music streaming services. First Monday, 27(1), 1–19. https://doi.org/10.5210/fm.v27i1.11783
Frenneaux, R. (2023). The rise of independent artists and the paradox of democratisation in the digital age: Challenges faced by music artists in the New Music Industry. DIY, Alternative Cultures & Society, 1(2), 125–137. https://doi.org/10.1177/27538702231174200
Hamilton, C. (2019). Popular Music, digital technologies and Data Analysis: New Methods and questions. Convergence: The International Journal of Research into New Media Technologies, 25(2), 225–240. https://doi.org/10.1177/1354856519831127
Hassan, J. (2021, November 21). Adele asked Spotify not to shuffle the tracks on her carefully curated … The Washington Post. https://www.washingtonpost.com/arts-entertainment/2021/11/21/adele-30-album-spotify-default-shuffle/
Kubala, B., Antoniak, M., & Szymkowiak, A. (2020). Music Sales and Artist Popularity on Social Media.
Mann, C. C. (2000, September 1). The heavenly jukebox. The Atlantic. https://www.theatlantic.com/magazine/archive/2000/09/the-heavenly-jukebox/305141/
Nwagwu, W. E., & Akintoye, A. (2023). Influence of social media on the uptake of emerging musicians and entertainment events. Information Development. https://doi.org/10.1177/02666669221151162
Rahimi, R. A., & Park, K.-H. (2020). A comparative study of internet architecture and applications of online music streaming services: The impact on the global music industry growth. 2020 8th International Conference on Information and Communication Technology (ICoICT). https://doi.org/10.1109/icoict49345.2020.9166225
Samanta, O. (2023, April 9). Spotify stats: Users, Revenue & Demographics 2023. Priori Data. https://prioridata.com/data/spotify-stats/
Sesigür, O. (2021). Playlisting, 44–49. https://doi.org/10.4324/9781003220541
Wang, A. X. (2018, November 1). An indie music expert explains why artists are turning away from record deals. Rolling Stone. https://www.rollingstone.com/pro/news/ditto-music-lee-parsons-interview-749510/
Wares, C. (2022, September 9). How social media burnout affects musicians. Berklee. https://www.berklee.edu/berklee-now/news/how-social-media-burnout-affects-musicians
Webster, J. (2023). The Promise of Personalisation: Exploring How Music Streaming Platforms Are Shaping the Performance of Class Identities and Distinction, 25, 2140–2162. https://doi.org/Ds:O//dIo: i1.o0r.g1/107.711/1747/611446144484281211002277863
Werner, A. (2020). Organizing music, organizing gender: Algorithmic culture and Spotify recommendations. Popular Communication, 18(1), 78–90. https://doi.org/10.1080/15405702.2020.1715980
cole manasco 