Selected Papers of #AoIR2022: The 23rd Annual Conference of the Association of Internet Researchers Dublin, Ireland / 2-5 Nov 2022 Suggested Citation (APA): Lastname, Firstinitial. (2022, November). Paper (or panel) title. Paper (or panel) presented at AoIR 2022: The 23rd Annual Conference of the Association of Internet Researchers. Dublin, Ireland: AoIR. Retrieved from http://spir.aoir.org. DELETING VIDEOS ON TIKTOK AS ALGORITHM RELATED IMPRESSION MANAGEMENT Daniel Klug Carnegie Mellon University Content Management and Strategies of Self-Presentation on TikTok Social media platforms allow users to create, share, and engage with social media content to communicate and network with others based on personal information (boyd, 2010). Self-presentation and impression management on social media is determined by individual and collective efforts alike (Hogan, 2010), users’ online perception is equally influenced by one’s own and others’ content (Walther et al., 2008). On the short-video app TikTok, users’ self-presentation is largely based on interaction and communication with and through user-generated video content rather than on profile information or through messaging (Chen et al., 2019). TikTok users likewise consume and produce videos, they are presented with individually curated video feeds based on their content engagement, and they present themselves through video portfolios on their content-centered profile page. TikTok users demonstrate algorithmic literacy, they are quite aware that and how content is customized (Oeldorf-Hirsch & Neubaum, 2021) and develop strategies and video creation practices (Klug et al., 2021). We find that such strategies extend to TikTok users’ profile management and result in the practices of deleting videos and making videos private as forms of impression management online. Social media users commonly delete their own posts because they no longer represent their thoughts, to protect their personal reputation (Statista, 2021), and because of mistakes or regret (Wang et al., 2011). Deleting content also involves unwanted messages posted by others (Madden et al., 2013). Untagging yourself from others’ content mostly happens due to conflicts around the message, for privacy reasons, or because of inappropriate representation of the self (Birnholtz, Burke, Steele, 2017). Our findings exemplify various levels of algorithmic literacy among TikTok users and illustrate how social media algorithms influence users’ impression management and their understanding of mediated realities. Suggested Citation (APA): Klug, D., (2022, November). Deleting Videos on Tiktok as Algorithm Related Impression Management. Paper presented at AoIR 2022: The 23rd Annual Conference of the Association of Internet Researchers. Dublin, Ireland: AoIR. Retrieved from http://spir.aoir.org. 2 Data Collection and Analysis In August 2020, we conducted 28 qualitative interviews with TikTok users in the U.S. Participants were recruited through location-based hashtags on TikTok (e.g. #pitt, #upenn#, philadelphia) and contacted via their linked Instagram profile. Participants were aged 18 to 25, had 300 to 450k followers, posted 15 to 425 videos, and had received 10k to 11.5m total likes. Questions concerned participants’ motivations, strategies, and practices before, while, and after creating TikTok videos. Interviews were analyzed following an open coding approach using a codebook to ensure validity of results through mutual agreement. Content-Related Self-Awareness: Deleting Videos We generally find that TikTok users deleted their videos because the content or message did not represent their views or attitudes anymore. Most participants described that they became aware of older videos or deliberately looked through their videos as part of profile maintenance, for example, deleting older videos because they felt their performance could have been better at the time. This shows TikTok users’ awareness for their self-presentation and for the perception of others through the content they posted. For some participants, this self-awareness extends beyond TikTok into offline reality. They explained how they deleted videos because they feared their TikTok content would negatively affect their impression in offline reality on people they know from TikTok. For example, participants were in retrospect embarrassed by content they posted and therefore set such videos to private or deleted videos if they already were private. Another main reason for deleting TikTok videos was lack of desired or intended feedback. Especially participants with larger numbers of followers had precise ideas of the views and likes they wanted to achieve with a video. They would closely monitor the immediate success of their video and often delete it within less than an hour if it failed to acquire the desired virality. In this way, users maintained the impression of a successful and popular profile consisting only of videos with a minimum level of popularity. Algorithm-Related Impression Management: Making Videos Private Moreover, we found impression management strategies related to participants’ assumptions of how the TikTok algorithm works and influenced their decision to set videos to private instead of deleting them. More experienced users explained that deleting videos appears suspicious to the TikTok algorithm and dubious to followers while private videos still add to one’s algorithmic portfolio and stay accessible for followers. In general, participants set unwanted or outdated videos to private because deleting would mess up their personalized algorithmic video feed and ultimately lead to missing the trends they are looking for. Moreover, participants felt deleting videos would affect the impression the TikTok algorithm has of their profile and eventually push their videos to the wrong audiences. To counteract, some participants would first create private videos to have the algorithm pick up sounds, hashtags, or filters they used and make the video public once one of these features started to trend. Few participants also kept videos set to private after they went viral to clean their profile but still benefit from this video’s success. 3 Conclusion Impression management strategies such as described are common for interaction and communication on social media platforms. Our results show that TikTok users create, evaluate, and maintain their online identity in regards to their experiences with and understanding of the app’s algorithm. We found TikTok users generally constantly evaluate their portfolio against current trends. They demonstrated certain levels of algorithmic literacy in deciding when and which content to delete or to set to private as a way of algorithmically influenced impression management. Our results show that TikTok users largely follow established strategies as part of impression management on the app, even if they only have small numbers of followers and likes or only occasionally post or interact with content. This provides new insights into the effort and thought that users put into making short content on TikTok and how they manage this content to create online identities. References Birnholtz, J., Burke, M., & Steele, A. (2017). Untagging on social media: Who untags, what do they untag, and why?. Computers in Human Behavior, 69, 166-173. boyd, d. (2010). Social network sites as networked publics: Affordances, dynamics, and implications. In A networked self (pp. 47-66). Routledge. Chen, Z. et al. (2019). A study on the characteristics of douyin short videos and implications for edge caching. In Proceedings of the ACM Turing Celebration Conference-China (pp. 1-6). Hogan, B. (2010). The presentation of self in the age of social media: Distinguishing performances and exhibitions online. Bulletin of Science, Technology & Society, 30(6), 377-386. Klug, D., Qin, Y., Evans, M., & Kaufman, G. (2021, June). Trick and please. A mixed- method study on user assumptions about the TikTok algorithm. In 13th ACM Web Science Conference 2021 (pp. 84-92). Madden, M. et al. (2013). Teens, social media, and privacy. Pew Research Center, 21(1055), 2-86. Oeldorf-Hirsch, A., & Neubaum, G. (2021). What Do We Know about Algorithmic Literacy? The Status Quo and a Research Agenda for a Growing Field. Statista (2021). Reasons for social media users in the United States to delete social media posts as of September 2018. URL: https://www.statista.com/statistics/935334/ adults-who-deleted-social-media-post-following-reasons-usa/ [February 27, 2022]. Walther, J. et al. (2008). The role of friends’ appearance and behavior on evaluations of individuals on Facebook: Are we known by the company we keep?. Human communication research, 34(1), 28-49. 4 Wang, Y. et al. (2011). “I regretted the minute I pressed share”. A qualitative study of regrets on Facebook. In Proceedings of the seventh symposium on usable privacy and security (pp. 1-16).