Selected Papers of #AoIR2024: The 25th Annual Conference of the Association of Internet Researchers Sheffield, UK / 30 Oct - 2 Nov 2024 Suggested Citation (APA): Annabell, T. & Rasmussen, N.V. (2024, October). An algorithmic event: The celebration and critique of Spotify Wrapped. Paper presented at AoIR2024: The 25th Annual Conference of the Association of Internet Researchers. Sheffield, UK: AoIR. Retrieved from http://spir.aoir.org. AN ALGORITHMIC EVENT: THE CELEBRATION AND CRITIQUE OF SPOTIFY WRAPPED Taylor Annabell Utrecht University Nina Vindum Rasmussen London School of Economics and Political Science Introduction The end of November marks Spotify Wrapped season: the time of year when Spotify nudges users to view and share aesthetically pleasing data stories ‘wrapped’ and repackaged from their listening behaviour. We approach Wrapped as an ‘algorithmic event’, defined as a moment in time in which there is a collective orientation towards a particular algorithmic system and associated data. Wrapped is an example of an algorithmic event in which people both celebrate and critique data capture, affording an opportunity to critically think through algorithmic systems. At the same time, Wrapped makes visible and topical the collection and aggregation of user data, which we investigate collaboratively through creative workshops. Theoretical framework The repackaging of user data that is central to Wrapped can be situated within the (re)configuration of the self through data assemblages. Cheney-Lippold (2011) refers to ‘algorithmic identities’ produced by algorithms that infer categories of identity based on data. Such categorisation of identity is projected onto individuals outside of their control. Prey (2018) coins the term ‘algorithmic individuation’ as ‘a dynamic socio-technical process engaged in enacting the individual’ (p. 1095). He notes the prevalence of algorithmic individuation on platforms like Spotify and how algorithms play an increasingly important role in subject formation by ‘reflecting back categorized images of our self’ (p. 1096). There has been limited academic research that examines these dynamics with Wrapped, which we argue are heightened through the re-presentation of data back to users. A notable exception is an analysis of #spotifywrapped tweets by Burgess et al. (2022) who refer to Wrapped as a ‘Spotify data selfie’. Our research is also informed by the analytical orientation towards understanding how people experience and interpret ‘algorithmic culture’ (Striphas, 2023), uncovering what it means for people that algorithms infer categories of identity and assemble narratives for them. For Bucher (2017), affective encounters with algorithms as part of everyday life involve what she refers to as ‘the algorithmic imaginary’. This concept emphasises the way that people perceive and make sense of their experiences with algorithms. Following this view, scholars have explored how people think about their ‘algorithmic identities’ and recommendations through arts-based methods and workshops (Lupton, 2021; Bishop & Kant, 2023; Siles et al., 2020), which we build on. Methodological approach To examine how people make sense of Wrapped as an algorithmic event, we have developed a workshop that enables users to unwrap layers of datafication. Drawing on ‘data feminist principles’ (D’Ignazio & Klein, 2020), participants engage in a series of exercises. The first one focuses on the 2022 ‘Listening Personalities’ or ‘Me in 2023 Characters’ cards from Wrapped. We suggest directions for discussion related to connections between music listening and taste, types of data collection implied by the categories and feelings towards labelling. The second exercise is a modified version of the walkthrough method (Light et al., 2018). Our version remains committed to the orientation to materiality, perceptions of affordances and built-in inequalities. In pairs, participants ‘walk through’ a particular Wrapped data story or the Spotify app, focusing on how data, personalisation and identity are (in)visible. As part of this process, they produce a diagram to record their observations and emerging insights (Figure 1). Drawing on craft-based approaches, the final exercise sees participants producing a physical artefact. Participants construct their own Wrapped that might rearticulate or critique Spotify’s version, using material objects such as CDs, vinyl, magazines, glitter and markers (Figure 2). We hosted nine workshops with more than 200 university students in the United Kingdom in 2023 and 2024 (see Annabell & Rasmussen, 2024). We obtained informed consent to audio record their contributions in the plenary discussions and photograph materials they produced. Although participants were invited to bring in Wrapped screenshots, we did not collect this as data. Instead, we focus on their interpretations of their own data. We position our participants as co-analysts, following the work of Robards and Lincoln (2017) and Markham (2021), and in our analysis highlight themes that arise from their contributions. Figure 1: Participants ‘walking through’ the interface Figure 2: Examples of creative artefacts produced by participants Preliminary findings In our early analysis of the plenary discussions, walkthrough diagrams and creative artefacts, we observe responses to the value of Wrapped that range from celebration to critical and resistant readings. At times, participants used Wrapped as a mnemonic vehicle to reflect on their patterns of listening as well as broader experiences with music. In such instances, we note how the claims that Spotify ‘knows you’ and the logic of Wrapped are reproduced, which was evident in the production of the creative artefact that echoed Spotify’s use of ‘top’ artists or songs to best represent Spotify listening and capitalise on parasocial relationships. Wrapped as revealing ‘truth’ about music consumption and taste was also contested. Some participants identified a friction between Wrapped and how they experienced Spotify. Wrapped was unable to convey the intricacies of experiences. For example, by restricting the user to one personality and collating user data, Spotify collapses diachronic, dynamic listening behaviour, which was considered by some as core to how their listening took place. This theme was continued in the production of artefacts that illustrated changes over time or interconnections between people, going beyond Spotify data stories. As part of critical reflections on Wrapped, participants also drew attention to tensions between claims of individuality and distinctiveness with an orientation towards collective patterns and genericity. We argue that Wrapped in these instances is experienced as personalised but not personal, repositioning the rhetoric of the platform through data stories. Finally, the construction of Wrapped as an algorithmic event was indicated by the way some participants discussed the lead-up to Wrapped season, which could be marked by feelings of trepidation and concern over what might be revealed. This anticipation was at times a catalyst for adapting their use of Spotify in an attempt to regain autonomy and influence the data collection that was assumed to underpin the data stories. In addition, the temporal dimension of the algorithmic event in which Wrapped becomes a marker in time, not only of time, was picked up in the production of some artefacts that played with its annual recurrence. Conclusion In this paper, we present emerging findings from our creative workshops in which we invite participants to analyse and discuss algorithmic events like Wrapped. In collaboration with participants, we tease out how normative assumptions are baked into Wrapped data stories and mobilise particular understandings of individuals, their habits, tastes and identities. We argue that algorithmic events like Wrapped are useful ways to think through data capture and algorithmic systems because of their familiarity and everydayness coupled with their popularisation and celebration of data. The phenomenon of ‘wrappification’ – by which we mean the repackaging of behavioural data that captures a particular activity throughout the year and the responses to the belief that we can ‘know’ ourselves in this way – speaks to such impact. The production of Wrapped iterations by other companies and individuals necessitates not only research but also opportunities for people to critically reflect on and discuss datafication and algorithmic identities. References Annabell, T., & Rasmussen, N. V. (2024, April 12). Spotify Wrapped: Workshop on repackaged music data [Video]. YouTube. https://youtu.be/Bu7lg9YzGj4 https://youtu.be/Bu7lg9YzGj4 Bishop, S., & Kant, T. (2023). Algorithmic autobiographies and fictions: A digital method. The Sociological Review, 71(5), 1012–1036. https://doi.org/10.1177/00380261221146403 Bucher, T. (2017). The algorithmic imaginary: exploring the ordinary affects of Facebook algorithms. Information, Communication & Society, 20(1), 30–44. https://doi.org/10.1080/1369118X.2016.1154086 Burgess, J., Albury, K., McCosker, A. & Wilken, R. (2022). Everyday data cultures. Polity Press. Cheney-Lippold, J. (2011). A new algorithmic identity: Soft biopolitics and the modulation of control. Theory, Culture & Society, 28(6), 164–181. https://doi.org/10.1177/0263276411424420 D’Ignazio, C. & Klein, L. (2020). Data feminism. MIT Press. Light, B., Burgess, J., & Duguay, S. (2018). The walkthrough method: An approach to the study of apps. New Media & Society, 20(3), 881–900. https://doi- org/10.1177/1461444816675438 Lupton, D. (2021). ‘Not the real me’: Social imaginaries of personal data profiling. Cultural Sociology, 15(1), 3–21. https://doi.org/10.1177/1749975520939779 Markham, A. (2021). The limits of the imaginary: Challenges to intervening in future speculations of memory, data, and algorithms. New Media & Society, 23(2), 382–405. https://doi-org/10.1177/1461444820929322 Prey, R. (2018). Nothing personal: Algorithmic individuation on music streaming platforms. Media, Culture & Society, 40(7), 1086–1100. https://doi.org/10.1177/0163443717745147 Robards, B., & Lincoln, S. (2017). Uncovering longitudinal life narratives: Scrolling back on Facebook. Qualitative Research, 17(6), 715–730. https://doi- org/10.1177/1468794117700707 Siles, I., Segura-Castillo, A., Solís, R., & Sancho, M. (2020). Folk theories of algorithmic recommendations on Spotify: Enacting data assemblages in the global South. Big Data & Society, 7(1), https://doi.org/10.1177/2053951720923377 Striphas, T. (2023). Algorithmic culture before the Internet. Columbia University Press. https://doi.org/10.1177/00380261221146403 https://doi.org/10.1080/1369118X.2016.1154086 https://doi.org/10.1177/0263276411424420 https://doi-org/10.1177/1461444816675438 https://doi-org/10.1177/1461444816675438 https://doi.org/10.1177/1749975520939779 https://doi-org/10.1177/1461444820929322 https://doi.org/10.1177/0163443717745147 https://doi-org/10.1177/1468794117700707 https://doi-org/10.1177/1468794117700707 https://doi.org/10.1177/2053951720923377