Module “ Recent Developments in Media Theory ” SoSe 2022 To which extent can everyday encounters of Facebook’s Ne ws Feed algorithm and TikTok’s For you page algorithm be compared? Term Paper Seminar: Algorithmic Cultures ( 17508 ) Module: Recent Developments in Media Theory ( 42034000 ) Lecturer: Dr. Vera Tollmann Date: June 9 th , 20 22 Sophia König 7th Semester Major Digital Media Table of Contents 1. Introduction ................................ ................................ ................................ ........................ 1 2. Theory ................................ ................................ ................................ ................................ 2 2.1 Facebook ................................ ................................ ................................ ................................ 2 2.2 TikTok ................................ ................................ ................................ ................................ ... 3 3. Analysis ................................ ................................ ................................ .............................. 3 4. Conclusion ................................ ................................ ................................ ......................... 8 Statutory Declaration ................................ ................................ ................................ ................. 9 Works Cited ................................ ................................ ................................ ............................... 9 1 1. Introduction Bucher (2020 , 1704) classifies Facebook’s News Feed as “one of the earliest examples of an algorithmically organized social media stream” . It was launched in 2006 as a single feed displaying content in reverse chronolog y and has since then developed into a fully curated stream of updates governed by machine - learning algorithm s Whilst Facebook has been around for almost 20 years, TikTok has surfaced as a social media platform in 20 16 The video sharing platform centr e s around an algorithmica lly curated stream of content. Since its emergence , it has been gaining users quicker than any other social media platform. In autumn of 2021, TikTok reported to have gained over a billion monthly active users. While the number of users is still lower than Facebook, Instagram and YouTube , none of these platforms have gained that many users in a short period of time (Brake 2022, 7) Both platforms present the user with an algorithmically curated feed of content Nevertheless, Facebook’s user growth especially within the younger user demographic has been slowing down whereas TikTok ’s community has been gaining more and more traction. T hus, t he main goal of this essay is to analyse and compare everyday encounters of the Facebook algorithm with the TikTok algorithm T his paper aims to uncover how the user experiences each algorithm and to which extent the user’s relation ship with the algorithm differs. In order to do so, the theoretical chapter will focu s on defining the framework of the social media platforms relevant for the analysis. The analysis includes a comparison of everyday encounters with the platforms’ algorithms . Here, the analysis will be structured according to themes that emerged in Taina Bucher’s (2018, 100f) interview study investiga ting user’s relations to the several algorithm s 2 2. Theory In the following chapter, the emergence and development of the subjects of analysis and comparison, Facebook and TikTok, are outlined. 2.1 Facebook Mark Zuckerberg launched the platform Facebook in the U.S. under the name “The Facebook“ in 2004 as a student community website (Hall 2021) Today, it is social media platform with the most active users. In 2021 Facebook reported to have over three billion users of which 1. 9 billion are active every day (n.a. 2022) On the platform, a user can create a profile and connect with others by “befriending” them. In addition, a user can follow companies, groups and other organizations. After befriending other users on the platform, users can direct message each other and post content such as media and text posts to their p rofile. If a user logs into their account, the home screen consists of their individual news feed displaying updates from friends and organi z ations Facebook’s news feed launched in 2006 and contains a stream of updates of friends and other interests (e.g. from liked organizations and companies). In the beginning, the feed was organized in reverse chronology and the newest updates were shown at the top of a user’s news feed. As one of the first social media platforms, Facebook restructured the news feed to be governed by machine - learning algorithms in 2009 (Newberry 2022) The company states that their news feed’s goal is to “ show everyone the right content at the right time so they don’t miss the stori es that are important to them” (n.a. 2014) The algorithm governing Facebook’s news feed has been subject of analysis for Taina Bucher in several of her publications. In 2012 , Bucher analyzed the effect the newly introduced news feed algorithm had on the content of her news feed as well as its functionalities in her publication “ Want to be on the top? Algorithmic power and the threat of invisibility on Facebook ” In her more recent publication, “ The right - time web: Theorizing the kairologic of algorithmic media ”, Bucher (20 20 ) examines how the news feed algorithm det e rmines the right - time to show content to the user. To do so, Bucher (2020, 1704 ) draws on the Facebook patent documents to analy z e how the algorithm works as well as an interview study conducted by her which focusses on how the algorithm is perceived and influenced by its users. The interview study was conducted within the framework of her publicat ion “ If...Then: Algorithmic Power and Politics ” (2018) It features user’s everyday encounters with the Facebook algorithm as well as their personal algorithm stories. In this paper the results of her interview study will be subject of analysis as to how users experience Facebook’s n ews f eed algorithm. 3 2.2 TikTok The video sharing platform TikTok was launched in 2016 by the Chinese company ByteDance (Brake 2022, 7) It became available worldwide after merging with the Chinese social media platform, Musical.ly, in 2018 and has since then gained a large number of users in a short period of time ( ibid. , 7) In April 2022, TikTok has registered over a bi llion active users with its largest user demographic in the U.S. spanning between 10 and 19 years old (Doyle, 2022) Once you open the app, a user has access to the main features of TikTok without having to sign up. The frontpage is made up of a “For you” - page. Here the user is presented with an algorithmically curated stream of videos. If a user wish es to follow creators or upload videos, TikTo k requires a sign - in to their platform. After signing up and following creators, a user gains access to two additional video streams. One stream is made up of content posted by accounts a user follows and the other newly introduced stream display videos po sted by “friends”. “Friends” are classified by the platform as users who follow you and vice - versa. In addition to viewing and uploading videos, a user can also direct message other users and live - stream content. In their publication “ Why’s Everyone on TikTok Now? The Algorithmized Self and the Future of Self - Making on social media ” Aparajita Bhandari and Sara Bimo (2022) examine users’ experiences and relationships with the TikTok algorithm By working with the walkthrough method of the app and conducting an interview study, the authors aimed to examine initial use and changing experiences with TikTok, regular usage patterns , types of content viewed and shared, participants’ likes and dislikes of the platform as well as forms of self - expression Here, the interview study will be subject of analysis as to how users experience the TikTok algorithm. 3. Analysis The analysis evolves around everyday encounters with algorithms of social media platforms, based on a comparison of the Facebook and TikTok algorithm . In order to systematically compare the encounters , the following analysis and comparison are structured a ccording to themes which emerged in Bucher’s (2018, 100f) interview study as to when and how algorithms came to matter for her participants . The analysis aims to examine user’s general awareness of the algorithm which Bucher (2018, 100) titled “ P eople aren’t a math problem” and “Enter the popularity contest” in her publication Furthermore, the analysis examines how users interact with the algorithm which Bucher titled “clicking consciously” in her publication 4 The analysis aims to uncover to which extent the encounters of the algorithms correlate or differentiate to understand how the user’s relationship with an algorithm ically curated content feed developed from one platform to the other. Awareness Participants of Bucher’s (2018) i nterview study observed Facebook’s news feed overall negatively. The accuracy of the targeted ads presented by the algorithm was perceived as creepy and it was regarded as negatively if the algorithm made assumptions about its users. One participant reported to have been algorithmically profiled after blogging about American singer Taylor Swift. After blogging about the singer, the interviewee received ads on Facebook related to the content of her blogpost (ibid., 101f) . The interview participant perceived the algorithmic mismatch as amusing first but then evaluated the targeted ads as “slightly o ffensive as they make assumptions about me, which I don’t like to think are true.” (ibid. , 101) Moreover, users labelled inaccuracies o f the algorithm as “intrusive” and “inescapable” (ibid. 103) . Overall, the Facebook algorithm attracted negative attention in the realm of the Bucher’s interview study. Users stated being frustrated with the algorithm i f it seemed to be deciding for them which content is presented to them and which is hidden One p articipant said: “I am feeling annoyed that the algorithm decides for me [...] it is so unpredictable and strange.” (ibid., 106) For the users, it is not clear why they are presented with certain content on their news feed whilst other content does not get shown. According to Bucher (2018, 108) , it makes a difference if the users set the parameters of their news feed themselves and the algorithm is merely “helpin g them along” or if the algorithm determines relevancy on its own without explicit input from the users. For the interview participants, the Facebook algorithm overstepped its boundaries and they do not feel in control of the content shown to them on their news feed (ibid., 108) TikTok users refer to the platform’s algorithm in a more positive manner. According to Bhandari and Bimo (2022 , 5 ) , users want the algorithm to “get to know them”. More specifically, the users are interested in the algorithm getting to know their personalities and interests as this would lead to a more accurate algorithmic experience on the platform One interviewee stated: “ The more time I spend on TikTok, the better it gets to know my personal likes or dislikes, and it gives me more and more content that I like.” (ibid., 5) Interview participants stated that the accuracy of the algorithm posed a significant draw to the plat form for them. In the conversations , users reported to possess an awareness that TikTok was collecting and sharing their personal data, however, this was seen as an “acceptable trade - off” 5 for the high - quality content shown on their “For - You” - page in return (ibid., 5) Furthermore, u sers praise the quality of the videos presented to them by the algorithm. One participant compared the recommended content they were presented by TikTok to recommended films to watch on Netflix (ibid., 7) . According to them , the Netflix recommendations weren’t as appealing as content curated by the TikTok algorithm. In comparison to experiences on other social media platforms, users said to be more consistently entertained on TikTok. On the one hand, participants praised the accuracy of TikTok’s algorithm. On the other hand, one participant also stated feeling unease if the algorithm was too accurate (ibid., 6f) Here, the interviewee referred to an experience where the presented conte nt reflected their physical appearance in addition to their sense of humour and interests. Further, some participants of the interview study viewed the algorithm as “restricting and overly fitting” after being exposed to algorithmically curated content on the platform (ibid., 6) Users on TikTok are being made aware from the onset that their experience on the platform is entirely shaped by the algorithm. Unlike the engagement with an algorithm on other pla tforms, users engage exclusively with the algorithm itself by primarily consuming content presented by the algorithm. Users do not set any parameters and engage in significantly fewer social interactions in comparison to other social media platforms. Moreover, Bhandari and Bimo (2022 , 6 ) observed users not feeling the need to follow certain creators to see more of their content. On TikTok, it sufficed to interact with the algorithm to gain access to similar content. Bhandari and Bimo (2022 , 5 ) evaluate this by saying that “the algorithm do es the work that ‘ following ’ does on other platforms, but more effectively and efficiently”. The manner in which users of Facebook and TikTok refer to their everyday encounters with the algorithm distinguishes itself greatly. A n accurate algorithm is perceived as disturbing and invasive by Facebook users whereas TikTok users aim for an accurate algorithm in exchange for high quality content. Subsequently, Facebook’s news feed al gorithm attracted negative attention if the user had the impression that the algorithm is deciding for them what content will be shown on their feed. The users felt out of control and evaluated the algorithm as encroaching. In contrast, users on TikTok primarily consumed content curated by the algorithm. The interviewees stated consciously engaging with the algorithm for it to get to know them and subsequently recommend “hyper - personalized” content. The algorithms of both pl atforms aim to present interesting content to the user. However, the user’s perception of the algorithm’s success varies greatly. The main difference between the platforms is the introduction of the algorithm. Many of today’s Facebook users registered on 6 t he platform before news feed was algorithmically generated. Further, users need to actively befriend fellow users and/or like pages to see updates on their news feed. Therefore, users set the parameters of the newsfeed themselves. In this realm, the algorithm is preferred to support the user’s interests and activity by recommen ding content and ads that align with them Subsequent ly, Facebook users expect the algorithm to occupy more of a background role in their news feed experience. On TikTok, the user engages with the algorithm in a more direct manner. After registering, the user is directly confronted with content curated by the algorithm. By interacting with the algorithm, the user has the opportunity to show the algorithm which content is interesting, and which is irrelevant. From the onset, it is cle ar to the user that their experience is entirely shaped by the algorithm. The algorithm occupies a central role on the platform . However, it needs to be noted that the user’s willingness to share information with the algorithm to increase its accuracy is h igher in TikTok in comparison to Facebook. Although one participant of Bhandari’s and Bimo’s interview study experienced unease if TikTok’s algorithm seemed to o accurate, the majority of participants praised the algorithm’s accuracy. In comparison, most of Bucher’s interviewees perceived Facebook’s news feed algorithm more critically if e.g., the targeted ads seemed too accurate. Subsequently, t he threshold for receiving recommendations f rom the TikTok algorithm is significantly lower in contrast to Facebook’s news feed algorithm. Interacting with the algorithm In her interview study, Bucher (2018) observed participants interacting with the algorithm to shape their experience on the platform especially if the content curated by the algorithm is not to their satisfaction. By doing so, the users acknowledge their influence on the algorithm. One of the participants described this action as “clicking consciously” : “I find myself ‘clickin g consciously’ every day. This is going to sound crazy, but I remember the one time I accidentally clicked on a former high school classmate’s profile and cursed under my breath: ‘damn it, now I’m gonna see her stuff 24/7.’” Users reported that this action gave them the impression of taking charge of their algorithmically generated news feed. By consciously interacting with content of which they want to see more of, participants aimed to influence the algorithms feedback loop in the hopes of seeing similar content on their news feed in the future. By engaging with the algorithm, the users aim to gain a sense of control over their newsfeed by redirecting the algorithm and not only let their content be auto generated by the algorithm itself. 7 According to Bhandari and Bimo (2022, 6), TikTok users have a high degree of “algorithmic engagement”. In their interview study, users were observed to know how to “work with” the algorithm to show them more relevant and enterta ining content. By doing so, users recognize that their relationship with the algorithm is dynamic and changeable. User s engage with the TikTok algorithm by commenting and liking videos as well as swiping away irrelevant videos to signal the algorithm which content is of interest to them . However, Bhandari and Bimo (2022, 5) did not observe users having the urge to follow creators they enjoy in order to see more of their content. Interviewees stated that the “mental energy devoted to the TikTo k algorithm exceeds that spent on algorithms of other sites (ibid., 6). Users refer to the algorithm with personifying and humanising language and the researchers observed the interviewees to have a close relationship with the algorith m (ibid., 5). In the paper the author s state: “Participants believed that they knew how exactly to interact with the affordances and activity corridors of the app to “work with” the algorithm so that it could provide them with more relevant or entertaining content.” (ibid., 6). Users engage greatly with the algorithm aiming to be presented with relevant and interesting content. However, the authors also note that the algorithm is not a “tool” with which the user can curate their own content. It is described to be largely “impenetrable” which exists to present con tent to the user reflective of their inner self (ibid., 6). Users o f both platforms interact consciously with the respective algorithm to improve the content presented to them . Yet, the attitude towards doing so differs. Whereas user s on Facebook report wanting to influence the algorithm to take charge of their news feed and control the content they want to see, TikTok users phrased it as “working with” with the algorithm to get it to present them with similar content. Users on both p latforms interact consciously with the algorithm to improve the content shown to them on their content feeds. However, the language with which users of both platforms describe their actions highlights the different perceptions of the algorithms. Facebook users describe it as a protective measure against a fully algorithmically curated news feed as opposed to TikTok users who are positively inclined towards the algorithm. In the presented interview study TikTok users interact consciously with the a lgorithm to improve its abilities to curate more interesting content. 8 4. Conclusion All in all, everyday encounters and perceptions with Facebook’s n ews f eed algorithm differ greatly with those of the TikTok algorithm. The analysis focused on examining users ’ general awareness of the respective algorithm as well as how users interact with the algorithm. Overall, the interview studies exemplify how Facebook users are more negatively inclined to the algorithmically curat ed news feed in contrast to TikTok users. Algorithmically curated and recommended content is viewed more critically by Facebook users as opposed to TikTok users. Here, the main difference between both subjects of analysis lies between their expected roles Whereas Facebook users expect the n ews f eed algorithm to occupy a supporting role to their existing n ews f eed, Tik Tok users are aware that their content is mainly curated by the algorithm. Facebook users are critically inclined to the n ews f eed algorithm if it oversteps its background role . Contrastingly, TikTok users are more open to algorithmically recommended content as the algorithm does not overstep its bounds in their view. In additio n, users of both platforms interact consciously with the algorithm to influence content presented to them. In the interview studies, however, TikTok users explain to do so to support the algorithmic activity whereas Facebook users aim to minimize and contr ol the influence the news feed algorithm has on their social media feed. The analysis has uncovered the importance of how an algorithmically curated content feed is introduced to the user. TikTok centres its platform around the user interacting solely with the algorithm whereas Facebook lets the user set its own parameters on their news feed. Therefore, Facebook users experience the algorithm as intrusive if the presented content is not to their liking. Contrasting, it is cle ar from the beginning for TikTok users that the algorithm is the main force in curating their content feed. Thus, users aim to have a positive relationship with the algorithm to get it to present them interesting content. To take this research further one can widen the comparison and analyse how the remaining theme in Bucher’s interview study titled “A lgorithmically I mag ina r y” is represented in both user groups. Moreover, it could also be of interest to investigate how each algorithm functions to possibly uncover if these findings offer possible explanations as to how each algorithm is e ncountered by the user. 9 25th May 2022 Statutory Declaration I declare t hat I have authored this thesis independently, that I have not used other than the declared sources / resources, and that I have explicitly marked all material which has been quoted either literally or by content from the used sources. __________________ __ ____ Date Signature Works Cited Bhandari, A., & Bimo , S. (2022). Why’s Everyone on TikTok Now? The Algorithmized Self and the Future of Self - Making on Social Media. Social Media+ Society, 8(1), 20563051221086241. Brake, M. (2022). Der Algorithmus, wo ich mitmuss? Fluter, 6 - 8 Bucher, T. (2012). Want to b e on the top? Algorithmic power and the threat of invisibility on Facebook. New media & society, 14(7), 1164 - 1180. Bucher, T. (2018). If... then: Algorithmic power and politics. Oxford University Press. Bucher, T. (2020). The right - time web: Theorizing t he kairologic of algorithmic media. N ew media & society, 22(9), 1699 - 1714. Doyle, B. (2022). TikTok Statistics – Updated April 2022. Wallaroo. Last accessed 22 nd May 2022 under https://wallaroomedia.com/blog/social - media/tiktok - statistics/ Hall, M. ( 2021). Facebook. Britannica. Last accessed on 25 th May 2022 under https://www.britannica.com/topic/Facebook. n.a. (2014). Showing More Timely Stories in News Feed. Meta. Last accessed on 22 nd May 2022 under https://www.facebook.com/formedia/blog/showing - more - timely - stories - in - news - feed n.a. (2022). Facebook gewinnt wieder mehr Nutzer. Süddeutsche Zeitung. Last acce ssed on 25 th May 2022 under https://www.sueddeutsche.de/wirtschaft/meta - facebook - quartalszahlen - 1.5574342. Newberry, C. (2022). How the Facebook Algorithm Works in 2022 and How to Make it Work for You. Hootsuite Blog. Last accessed 22 nd May 2022 under https://blog.hootsuite.com/facebook - algorithm/#:~:text=2009%3A%20Facebook%20premieres%20its%20first,be%20prioritized %20in %20their%20feed.