On following and becoming a follower: A phenomenological analysis

This paper uses phenomenological analysis to consider the phenomenon of becoming a follower. The data includes my field notes from mediated encounters with three of the most-subscribed YouTube content creators, as well as my personal recollections about these experiences, which are interspersed as event memories throughout my notes. Analyzing my event memories of these encounters enables me to follow my affective relationship to these YouTubers and their content, to untangle some of these knobby multiplicities and map the possibilities of following as they emerge. As I follow following around, as a researcher and (perhaps) as a follower myself, I reflect on the role of algorithms shaping who, what, and how we follow online media. Lastly, the methodological ramifications of a researcher becoming a follower – how following and researching bleed into each other, and why this matters – are also explored in this paper.


Introduction
What does it mean to follow, to be a follower?Someone who trails behind others, who follows people or commands, a supporter, or a devotee -perhaps blindly.The action of following has various connotations: to follow somebody on the street might be threatening; to follow a religion demonstrates faith (perhaps also blindly).On social media, following another person grants you access to their (online) life, through a stream of posts and comments that may (not) be

Following YouTube
YouTube is the second most-visited website and social media platform in the world as of 2023 (Dixon, 2023) with over 246 million monthly active users in the United States alone (Ceci, 2023).The platform was founded in 2005 to encourage 'ordinary' people to create content and communities, with the mentality that 'anyone can participate,' connect, and interact using the platform (Burgess & Green, 2018, p. 123).This participatory approach is central to YouTube's business model, driving the site's content, personalities, and metrics (Burgess & Green, 2018, p. 124).However, the platform has become institutionalized and now highlights professional rather than amateur content (Kim, 2012).The most successful YouTubers have leveraged their following to negotiate sponsorships, headline tours, and produce merchandise (Cunningham & Craig, 2017).The YouTubers in this study are professionalized and have used their central channel as an anchor for numerous brand extension opportunities.
These YouTubers are considered microcelebrities because they use social media to connect with their followers and create commercialized content based on their online persona (Abidin, 2015;Berryman & Kavka, 2017).Microcelebrity is not just a status of online celebrity but also the set of practices that one uses to attain that celebrity; chief among these practices is the use of "strategic intimacy," in which online users selectively display their emotions and personal lives (Marwick, 2015;Raun, 2018b).Such strategies are a form of affective labor, enabling them to produce, manage, and circulate affect in an open-ended social relation (Carah, 2017;Dowling, 2007;Duffy & Wissinger, 2007).This labor is deployed to enhance perceptions of audience connection (Abidin, 2015) and authenticity (Cunningham and Craig, 2017), which is in turn mobilized to influence the beliefs and practices of followers (Hearn, 2010).The term "influencer," which is typically used to refer to these digital microcelebrities, reflects the commercial logics at work in the development and goals of their personal brand (Banet-Weiser, 2012;Marwick, 2013).As influencers, the YouTubers in this study carefully balance commercial and affective logics to grow their following, which is to say, grow their business.Analyzing the influencer-follower dynamic from the researcher-follower perspective enables closer consideration of how this balance is (not) achieved.
Vlogs, or video blogs, are one of the most popular forms of content on YouTube, in part because they promote perceptions of authenticity, intimacy, and community among viewers (Cunningham & Craig, 2017;Horak, 2014;Raun, 2018b).YouTube vloggers share their everyday lives, experiences, and thoughts with viewers, using storytelling to construct a (branded) identity (Berryman & Kavka, 2017;Jerslev, 2016;Tolson, 2010).Jerslev (2016) defines "YouTubers" as vloggers who post content on their personal channels, noting that the platform's emphasis on participation and connection influences their performances of intimacy.Raun (2018b) also notes that YouTube vlogs evoke and capitalize on intimacy as a genre through their camera work, onscreen emotional disclosure, and interactions with viewers.The YouTubers in this study use these gestures of intimacy while also positioning themselves as more traditional celebrities, with all the luxurious trappings of fame and wealth: expensive cars, lavish trips, designer clothing.Marwick (2015, p. 156) argues that this aspirational content (re)positions followers as fans, an "adoring audience" that is interested in gawking at the lives of these users as if they were a more traditional celebrity figure.I reflect on YouTube's culture and affordances throughout this study to consider how they influence relations among creators and followers on the platform.

Following algorithmic media
Algorithms are machine-learning programs that use data to predict future online behaviors (Gillespie, 2014).Social media platforms such as YouTube use algorithms to curate, recommend, and moderate content as well as advertising (Saurwein & Spencer-Smith, 2021).Algorithmic data is presented as value-neutral but retains the biases of its human creators, shaping not only the viewing patterns of online users but also their visibility and humanity (Benjamin, 2019).Social media platforms manage not only the visibility of users, but also that of their algorithms (Flyverbom, 2016).YouTube's algorithm is similarly a "black box": a closed system considered proprietary knowledge and therefore known only to the company itself (Pasquale, 2015).Previous work has considered how users produce and manage content in the face of this opacity based on their "algorithmic imaginary" (Bucher, 2018;Van der Nagel, 2018), including influencers (Cotter, 2019) and brands (Carah, 2017).I use a phenomenological lens to consider how my own algorithmic imaginary structures my experience of researching-following, though it must also be noted that the content I encounter is also shaped by each YouTuber's understanding of the platform's algorithm and their desire to boost their visibility (Cotter, 2019).
Researchers have reverse-engineered YouTube's algorithm to explore its use of collaborative filtering analysis, which calculates predictions based on users' personal data-such as their age, geographic location, gender, and language preferences-but also aggregated viewing practices across the site (Airoldi et al., 2016;Arthurs et al., 2018).The YouTube algorithm also uses "signals" such as video clicks, watch time, sharing, and ratings to tailor recommendations to "content that uniquely inspires, teaches, and entertains" (Goodrow, 2021).However, the algorithm has simultaneously facilitated the rise of extremist content (Roose, 2019;O'Callaghan et al., 2015) and gendered, classed hierarchies (Bishop, 2018) on the platform, as well as the suppression of content that is not considered "advertiser friendly" (Caplan & Gillespie, 2023).A growing body of qualitative research has analyzed the role of algorithmic data in shaping user interactions with online media (e.g.Schmidt et al., 2019;Lupton & Watson, 2020;Thorndahl & Frandsen, 2021).Like Are (2021) I mine my own experiences on social media to consider what I (do not) follow on YouTube, and how this is mediated by algorithms.

Becoming a researcher-follower of YouTube
My phenomenological inquiry analyzes my lived experience of watching these YouTubers and becoming a follower, but it is important to note that this is not the central focus of this inquirynor is it the focus of phenomenology.Van Manen (2017) explains that in the phenomenological method, "what appears in consciousness is the phenomenon or event that gives itself in lived experience" (p.811), rather than our interpretation of that experience.Phenomenology is therefore the study of "the experience as lived" and "how phenomena are given to use in consciousness" (p.813, my emphasis).By "lifting" and "holding" an experience with our "phenomenological gaze," we are able to make "an ordinary lived experience…quite extraordinary" (p.812).This gaze is attuned to the structural meanings of the phenomenon, and to describe and interpret them in depth (Van Manen, 2014).Put more simply, rather than asking what something means, phenomenology asks, "what is this phenomena like?"This enables the researcher to make this singular experience knowable (Van Manen, 2017).As a philosophy, phenomenology therefore reflects an onto-epistemological perspective that lived experience, or "being in the world," can be interpreted and produce knowledge (Dowling, 2007).I also draw from the work of Sara Ahmed (2010) to follow the sensation of becoming a follower around.In The Promise of Happiness (2010), Ahmed follows happiness around to explore "what histories are evoked by the mobility of this word" and its presence or absence; she "goes where it goes," and in the process notices "what it is up to, where it goes, [and] who or what it gets associated with" (p.14).As she follows this concept around, she examines refractions of happiness and so-called "happy objects" in her own life and those of others, through literature and film, theory and history, to consider how happiness shapes our values and "makes certain objects proximate," and in so doing "shape how the world coheres" (pp.14-15).Ahmed's history of (un)happiness is genealogical: it offers "different readings of its intellectual history [and considers] who is banished from it" (p.17) from her perspective as an outsider within (Collins, 1986), that is to say, as a queer woman of color.She uses phenomenological inquiry to follow happiness, to "unravel" the concept and its appeal (Ahmed, 2010, p. 18).By doing so, Ahmed's work asks how this phenomena is lived, experienced, and productive: what it is happiness like, how does this feeling shape our world(view), whose feelings shape this concept?I follow following around to consider how it feels to be a follower, how it shapes our (on-and offline) world, and, through brief reflection on algorithmic engagement, whose feelings shape this concept.Like Ahmed, I am not trying to "find" following so much as I am (re)producing my own experience of becoming a follower and following following around, by mapping it as it emerges alongside my data (p.18).In doing so, I reflect on YouTube as an (un)happy object, which we might follow in order to follow (un)happiness (p.21).This involves reckoning with how YouTube content is produced, and how it acquires the shape of the bodies (human and otherwise) that inhabit this space (Ahmed, 2007, p. 156).It also requires a closer look at the affective potential of YouTube(rs).While affect is often used interchangeably with "emotion," Massumi (1995) refers to it as "intensity": he argues that affect is autonomous, a "synaesthetic" state beyond cognition, language or even the body, even as it is often expressed or experienced in those forms.Affect is intensity, but it is also a force (even if it is not always forceful) "that can serve to drive us toward movement, toward thought and extension" (Gregg & Seigworth, 2010, pp. 1-2).Previous scholarship has considered how social media platforms channel affective flows to increase engagement (e.g.Carah & Dobson, 2016), and how influencers calibrate their content to manage affective responses among their followers (e.g.Abidin, 2016;Raun, 2018).I explore how feelings of relating, knowing, and trusting drove my research on social media influencers in this study.
Following Ahmed (2004) and Paasonen (2016), I therefore conceptualize the YouTube vlogs in this study as "sticky tapestries of value," which generate and sustain affective patterns of relation among YouTubers and followers through algorithmic intensification (Papacharissi, 2015).The YouTube algorithm, which aims to "delight" users by shaping what they follow based on what they might like, is an object with feeling that creates affective impressions among YouTube users (Ahmed, 2010;"How YouTube Works," n.d.).These impressions are sticky in part because they are layered and often ambivalent (Paasonen, 2016).For example, Roose (2019) considers how YouTube content can direct users into unhappiness and alienation, and yet they continue to seek new content: they remain locked in a relation of cruel optimism, drawn to "compromised conditions of possibility" (Berlant, 2013, p. 24).Hallihan and Striphas (2014), Carah (2017), and Saurwein and Spencer-Smith (2021) also reflect on the relationship between users, platforms, brands, and algorithms, which promise choice and control even as they increasingly constrain users' agency and data security.Following, in other words, is not necessarily happy; it is often performative, though not always hopeful or even hap-ful (Ahmed, 2010).Ahmed's (2010) and Berlant's (2013) work, and perhaps also this paper, offer a glimpse of how objects such as YouTube not only facilitate but promise affective engagement.

Methodology
This paper adapts Sinclair's (2020) method of "event memory" to analyze my notes from several media encounters with three YouTubers.This enables me to use my own narrative reflection, about my experience of these encounters as well as my methodological process, to "trouble the flatness of the transcript, to open other possible entryways" (p.451) into the notion of following.The sprawl of subjects and data in this search are not unlike the "knobby," rhizomatic shoots of crabgrass that Sinclair describes in her article: by using event memory, I'm able to "construct [a] place in which to get lost, a place to wo(a)nder, a place of multiplicity" and messy truths, not only about the language of "following" but also my own relationship to the concept as a researcher and user of social media (p.449).In other words, I'm able to follow my affective relationship to these YouTubers and their content, to untangle some of these knobby multiplicities and map the possibilities of following as they emerge.This approach opens up my field notes and provides space(s) to wo(a)nder, to disrupt the hierarchy and linearity of narrative as well as transcription: it brings the echoes of my memory onto the page (p.450-451).These reflections enable a more dynamic expression of my lived experience as a follower-researcher.

Data for analysis
This study is drawn from a larger project analyzing fan merchandise produced by three prominent YouTube vloggers: Jake Paul, David Dobrik, and The Ace Family.Paul was originally known for pranks, stunts, and music videos, which he has parlayed into a second career as a professional boxer.Dobrik's vlogs also include extravagant pranks and stunts, often filmed with a group of friends and fellow YouTubers known as "The Vlog Squad."The Ace Family follows the lives of the five titular family members-Austin and Catherine McBroom and their three young childrenthrough everyday activities, pranks, and viral challenges.I selected vloggers because these microcelebrities and their content are endemic to YouTube (Burgess & Green, 2018;Jerslev, 2016).I selected these creators due to their popularity during initial data collection at the inception of the overall project in early 2020: at the time, they were three of the most-subscribed accounts in the United States in the People & Blogs category, according to the social media analytics firm SocialBlade.Each of these vloggers has also founded multiple ventures beyond YouTube, including merchandise, media, and lifestyle brands, and therefore offered a wide variety of transmedia material for analysis.Finally, all three produce content in English, my native (and sole) language.
Data generation I watched and analyzed 20 videos from each creator from January-August 2020, which included 5 of their most popular videos and 15 of their most recent videos.I took detailed field notes about video content, themes, and video metrics, including the date posted, the video summary, and the number of video likes and comments.I also documented the 20 most popular viewer comments.These comments are included in analysis because they were posted in a public forum with the expectation of visibility and engagement due to YouTube's system of "upvotes" in the video comments.However, I anonymize the commenters' usernames to preserve their privacy beyond this forum and thereby minimize potential harm (Franzke et al., 2020).
After this initial data collection ended, I attended a virtual interview with David Dobrik.I include this interview in this analysis because it was my first and most memorable encounter with the feeling of being a follower.The interview took place on Zoom several months into the COVID-19 pandemic, which halted Dobrik's weekly posts that I had just begun watching for my research project.When I logged onto the Zoom call and the interview began, I smiled unconsciously; when Dobrik turned his camera to show fellow Vlog Squad member Jason Nash, I felt a small shot of recognition: I knew him, too!I turned over this feeling of becoming a follower-of following content and people, and the feeling of knowing these people that I experienced.As I continued to analyze my field notes from the other YouTubers in the project, The Ace Family and Jake Paul, I noticed similar sensations of recognition, as well as irritation and interest: in other words, affective responses to their content.I therefore (re)turn to my field notes from the most popular video for each YouTuber in this paper to explore how my event memories shaped my data and analysis.The Ace Family video, "THE BEST PROPOSAL OF ALL TIME!!! (JUMPING OUT OF A PLANE)" (August 26, 2017), documents Austin McBroom's elaborate proposal to his now-wife Catherine, which involved the couple skydiving and later riding off in a horse drawn carriage.The Jake Paul video, "It's Everyday Bro (Song) feat.Team 10 (Official Music Video)" (May 30, 2017), is a music video for a novelty rap song.I selected these videos for two reasons.One, they have the most views of their content, and are therefore followed by many people.My data collection, which included comments on each video, also indicates that some followers return to old content to watch it again: following, in other words, does not suggest linear consumption.Selecting these highly followed videos therefore enables me to focus not only on my experience of following, but to also consider how other viewers follow these YouTubers.I also selected those field notes because they were compiled toward the end of data collection for each subject, at which point I had watched their videos multiple times: I had learned about their lives, could anticipate their catchphrases and behaviors, and had even begun to laugh at their antics despite myself.I had gone from researcher to follower-but how?

Data analysis
To consider this question, I wrote my event memories into my original Word document of field notes, then bolded some words and phrases to indicate my initial thoughts and meaning units.I then completed a memo in which I reflected on the process of writing my event memories and pondered some methodological choices and progressions.After a few days, I returned to my notes and printed them out so that I could further annotate the notes and memories with handwritten interruptions: circling words, drawing lines, jotting down questions and provocations.The temporal distance and manual interaction helped me make connections and establish more concrete meaning units.Like Sinclair (2020, p. 452), each of these reading and questioning sessions within the transcript helped me to indicate my "positions and assumptions," as well as my interpretations, so that they "can become explicitly a part of [my] data."This reflection enables further insight into what it is like to become a follower and follow following around, and how the knobby shoots of "following" appear within the spaces of my data and analysis.

On following and becoming a follower…
Reconceptualizing Educational Research Methodology 2023, 14(1)

29
The question of how social media data is produced, followed, and analyzed is of significant interest given the widespread use of algorithmic surveillance and information mining in these spaces (Benjamin, 2019;Bucher, 2018).Digital autoethnography has been a useful tool for researchers to reflexively interrogate social media use (Atay, 2020) and make algorithmic oppression more visible (Brown, 2019).Scholars have also used an autoethnographic lens to analyze the algorithmic suppression of online content (Are, 2021), the relational production of data on learning platforms (Lange, 2021), and mourning on social media (Raun, 2018a).Others have utilized posthuman and new materialist theory to explore human-data assemblages (Lupton, 2019;Markham, 2021), with specific attention to the ways in which the researcher is entangled with and co-produces these assemblages (Lupton & Watson, 2019).This study similarly mines my personal data but uses phenomenological inquiry to explore "the multiplicity of the experience" of conducting research with(in) an algorithmic feedback loop (Papacharissi, 2015;Sinclair, 2020, p. 452).I fold my researcher-follower experience into the data to make the interpretations and assumptions behind these interwoven phenomena more visible.
Before I delve into my analysis, it is important to address my positionality.My YouTube homepage and suggested videos are specific to my geographic location; gender; race; sexuality; language preferences, and videos that I have previously watched, among other factors (Goodrow, 2021).In other words, as a White, cisgender, queer woman who speaks English and lives in the southeastern United States, the content that I encounter is tailored to my perceived subjectivity.Furthermore, as I (re)watched videos from these creators, their content increasingly appeared on YouTube and across other social media platforms, an algorithmic convergence that made it difficult to separate my research and more casual online activity.This study is therefore conducted from within the "speculative middle" (Springgay & Truman, 2018, p. 206) of an algorithmic feedback loop of my own making (Papacharissi, 2015).My use and exploration of event memories is supplemented by the use of memos to reflect on inconsistencies, anxieties, and omissions in my analysis due to my positionality as a researcher-follower of algorithmic media (Clarke, 2005).These memos also make my everyday navigation of algorithmic media (more) visible to promote (more) critical and self-aware analysis of this digital platform (Risi & Pronzato, 2022).

Analysis
As I circled, underlined, and drew arrows around my printed data to reflect on what it is like to become a follower, relating, knowing, and trusting emerged as key meaning units and are explored in this section.I also used my event memories to map my "intellectual web"-that is, the interconnected ideas, terminology, and texts that I bring to my research-as I followed the experience of following following around (Sinclair, 2020, p. 450).This is a rather grandiose way of saying that I was also exploring how it feels to conduct this research on following, to consider how my data collection and analysis make relating, knowing, and trusting (im)possible.

Relating
Throughout my event memories I found myself relating to these YouTubers and their content.Sometimes that felt like connecting with a YouTuber, such as when was I drawn into the Dobrik interview-particularly during an exchange with the hosts about tea preferences and habits, during which my emphatic "SAME" and "WHAT" interrupted my rather "clinical" field notes, which until that point had attempted to bracket.Like Dobrik, I note in my event memories that "I have very strong tea opinions," and when he stated his own I felt like I could relate to him more than in my previous encounters with his content.Relating, then, felt like turning a shared preference into a connection, like developing a relationship.YouTubers actively relate to followers in their videos to develop these relationships, which can make them more trustworthy and therefore highly effective advertisers (Chapple & Cownie, 2017; Lee & Watkins, 2016).Dobrik adeptly drew the hosts and me into his home (physical: his residence; emotional: his channel or storyworld) during his interview by communicating directly and intimately with the hosts, and working to relate to them (Abidin, 2015).In the Ace Family video, Austin McBroom makes eye contact with the camera and changes the tone of his voice, sometimes whispering to the viewer and bringing the camera closer to his face to draw them into his plans for his proposal.Such moments of calibrated intimacy demonstrate a sophisticated use of camera work and affective labor to engage followers (Abidin, 2016;Raun, 2018b).These YouTubers' ability to relate to viewers helps build relationships, turning casual viewers into people who want to return, and reciprocate: to follow.
Following this data has made me reconsider what kind of relating builds relationships, and how the tone of those relationships evolves over time.Though Jake Paul's music video steers clear of connection or communication, its high level of engagement-over 287 million views and 1.2 million comments at the time of writing-suggests that many people are still relating to this content 4 years after its release.However, the tone, phrasing, and timing of the vast majority of these comments indicate that follower relationships can not only be unfavorable, but perhaps even compulsive: Sometimes I feel useless, but then I remember this video has a like button (JP Commenter 1, October 2021).
Whenever I feel sad or depressed about my life, I just watch this to help remind me that it can always be worse (JP Commenter 2, June 2021).
"Two months ago you didn't know my name" / Me: dude I still don't know who you are (JP Commenter 4, 2019).
In my event memories, I wo(a)ndered through these comments and pondered: "what is the value of this [video], what purpose does it serve to JP and his followers?"These commenters appear to have returned to the video, perhaps repeatedly, over time, though whether they actually watch the video is unclear.These anti-fans may not like the text, but nevertheless care deeply about it and wish to share their opinions about its cultural impact (Gray, 2003).Despite these repeated viewings, their derisive commentary suggests that they are not engaging with the video because they like it; at the same time, many comments have an almost gleeful tone.These anti-fans appear to experience pleasure in sharing their opinions with others, emphasizing that "dislike is as potentially powerful an emotion and reaction as is like" (p.73).Finally, many of these comments had accumulated thousands of likes themselves: followers, then, may use this form to relate not only with the YouTuber, but to form an "anti-fan community" with other followers (Gray, 2005).Anti-fandom literature offers insight into how following creates other followings: how it might be reluctant, yet strangely compulsive and even pleasurable (Click, 2019;Gray, 2005;Herman & Jones, 2013).Or perhaps it is not about whether this content is pleasurable-after all, objects can be pleasurable even if they are not pleasing (Ahmed, 2010)-rather, that the content captures the viewer's attention, generating value from their affective engagement (Cote & Pybus, 2007).Put differently: these moments of (not) relating are calibrated to "engage individuals in uninterrupted flows of attention and data sharing," to optimize and orchestrate affective action for YouTube(er)'s profit (Carah, 2017, p. 398).
My notes and event memories for the Ace Family and Jake Paul videos also show moments of (guilty) pleasure.This includes liberal amounts of what I labeled "snark," small asides (some overt) in which I mocked or questioned the creators or their content.Much of this was based on my knowledge of influencer behavior and social media networks, which has been cultivated as a researcher but also as an active internet user.Andrejevic (2008) considers how sarcastic anti-fan commentary is often wielded to demonstrate an "insider's perspective" of media texts, an expression of fan behavior and taste via critique.Herman and Jones (2013) argue that snarky critique enforces particular taste cultures, which can impact academic studies of fandom as well as the fandom itself.I wrestle with this in my event memories: "How do my thoughts/feelings/notes/analysis [change] as I follow"?Relating and reacting to these YouTubers changed the way that I related their video content: what I recorded, what was left out, and how I frame it, even (and especially) in this paper.It is crucial to remain attuned to these "wanted and unwanted feelings" so as to use them "as an ethical resource" (Ahmed, 2010, p. 216).
My event memories therefore provide a tangible record of how I relate to these YouTubers, making my resulting interpretations visible in my field notes and analysis (Sinclair, 2020).They also further complicate a purely positive understanding of relatability.As I follow Austin McBroom's documentation of his proposal, I contemplate how it feels "to live so much of their lives in front of other people" when he talks about having several cameras present; I "cringe" when he gives a ring to his infant daughter as a promise to "be here for you, no matter what" on her left hand, while leaving her right hand clear; I reflect on my memories of skydiving, which makes me question their clothing choices and then reconsider whether I trust that the proposal was a surprise.Relating to these YouTubers therefore did not always feel positive or friendly; I often sought points of contrast to emphasize our differences and critique their choices.Mapping my event memories made me realize that I follow certain moments more (or less) closely because I am able (or not) to map my experiences onto certain YouTubers or video narratives.Being "relatable" indicates that an influencer is able to show a persona or experience that is recognizable, that can be seen and mapped.Being a follower feels like mapping your experiences onto others, like using your experiences as a map of their motives.

Knowing
But what of these YouTubers' motives can be seen or known, by their followers or researchers such as myself?The second meaning unit, knowing, emerged from moments in my field notes and event memories where I asserted or questioned my knowledge of these YouTubers.More specifically, my field notes often revealed moments of my "positions and assumptions" about social media influencers, or YouTube itself (Sinclair, 2020).These moments of knowing were only visible when I untangled my event memories of the data.For example, I recognize the car that Austin and Catherine drive in their video: I note that it "appears to be the infamous Shelby," which is the name they gave their Jeep.They sell Shelby in a video that I watched as part of my merchandise project; originally titled "TODAY IS A SAD DAY FOR US…" (February 6, 2020), the video follows the couple as they prepare and take their car to a used car dealership.That video includes quite a few critical comments about the Ace Family's privilege and motives, such as, "I'm so sorry for your loss.I know it'll take you guys a while to get over such a huge loss, but you'll get through it.Having just 5 cars left in the garage is just unacceptable, my condolences to you guys" (Ace Commenter 1, December 2020).As I continued to follow the Ace Family in this data, I saw many more comments such as these on their videos, and encountered some articles and reaction videos from other YouTubers about their suspicious business practices.Recognizing and referring to the car as "the infamous Shelby" makes me feel like I know what has happened in these videos, which makes me feel like I know who Austin and Catherine are.My snarky tone also suggests that my "insider's perspective" as a follower and researcher has given me increased insight into their channel and personas (Andrejevic, 2008).Yet I continually wrestle with what I know, and indeed what can be known, about each of these YouTubers throughout my event memories.
When annotating my field notes from Jake Paul's music video, I grappled with the realization that my knowledge of his business practices and familiarity with his persona had impacted the way that I was documenting his content.I followed some of Paul's exploits prior to conducting this research: I knew about his relentless merchandise promotion (Stokel-Walker, 2018) and his questionable online education businesses (Leskin, 2020), and that his brother Logan had filmed a suicide victim (BBC Asia, 2018).However passively I had followed Jake Paul, my feelings about him coalesced into a sensation that I knew him, which followed me as I watched and analyzed his content.My knowledge of Paul's tactics, as well as my interest in fashion and fan merchandise, also directed me toward following the clothing in this video more closely.The video features quite a bit of Jake Paul merchandise, situated alongside popular streetwear brands such as Adidas and Thrasher: I note in my event memories that "these videos definitely seem like merch promotions."Following this promotional activity felt like vindication that my perception of him was accurate: that I really did know Jake Paul.Sometimes following is active and leads you exactly where you want to go.
Like many of Paul's followers, I had sought to indulge my (dis)pleasure in his persona."Knowing" him is therefore an epistemological but also methodological quandary: how do I know what a YouTuber is like, and how does what I like about them impact my data collection and analysis?This coalesced most acutely in my notes on Paul.Towards the end of the video, I write in my event memories that "I really, really don't like JP" due to his "aggressively self-aggrandizing behavior."In hindsight, this behavior might have been more present because of my feelings of knowing.I am still working through the epistemological aspect of this meaning unit, and what it means to know a YouTuber.Though I acknowledge that Paul's rapper persona is a "role," it feels like an alter ego: I know that he can't really be known, but I still feel like I know enough of him to know when he is On following and becoming a follower… Reconceptualizing Educational Research Methodology 2023, 14(1) 33 playing a character.Perhaps knowing a YouTuber is slippery, always evading your grasp-but following feels like wanting to reach out and grab them again, to see if you can pin them down this time, to finally see who they really are.

Trusting
The last meaning unit, trusting, became evident as I moved through the data.When annotating my notes from The Ace Family's video, my field notes almost immediately indicated my skepticism that the family and their content were inauthentic.Interspersed throughout the notes for this video, as well as the Jake Paul video I analyzed, were moments of mocking or dismissive language, plus a number of parenthetical asides that questioned whether the video content or creators were genuine or "real."I wrestled with my perceptions of authenticity on YouTube via analysis of David Dobrik's interview, noting that moments of feeling "realness" seemed to be moments of becoming a follower; becoming a follower felt like feeling someone is real.However, this expanded data set introduced a related concern: how feeling "realness" (or not) impacted how much I trusted creators and their content.As I lifted and held these feelings of (dis)trust within my phenomenological gaze (Van Manen, 2014), my experience of being a follower was less about what was authentic or not; when watching these videos and taking notes, I was working through my feelings about each creator to discern whether or not I could trust them to show something that was authentic to their persona or narrative, which I had cultivated based on my understanding of their personality or content.I can follow this shoot of (dis)trusting throughout my field notes and event memories (Sinclair, 2020).When reviewing my notes from the Jake Paul video, I reflected that it felt like "an ad for his merch line more than anything."Paul wears an "Ohio Fried Chicken" t-shirt, a design that was created for and then promoted in a previous music video of the same name.He also makes several references to his merchandise in the lyrics for "It's Everyday Bro," such as "always plug, merch link in bio" (3:19).Following Jake Paul in his video content, media appearances, and press made me particularly sensitive to his relentless production and promotion of self-branded merchandise (Stokel-Walker, 2018).Following someone within and outside of social media can lead you to distrust.
But trust, like following, is not a linear and unilateral process.When I revisited my event memories of the Dobrik interview, there are several moments where I note that I "wanted" something to be real, even when I acknowledged that it was most likely staged.I get "caught up in the moment" of the interview and describe feeling like Dobrik is "being genuine" due to his body language, speech, and appearance.However, I also note that while he's not wearing his own merchandise, he promotes his podcast and perfume line, and one of his friends is wearing David's Vlog merchandise during the interview.My event memories try to untangle this question of trust: [His language makes it seem like he's] In it for the craft and to have fun, which makes me feel like he's being genuine, not like he's trying to sell me things.(But he is selling things and making a lot of money: how do I reconcile this, how does he?How do other viewers?)Do viewers feel like they need to reconcile [this], or am I just overthinking?
My ambivalent affective impressions of these YouTubers "shifts and refocuses" my attention to these YouTubers' potential motivations (Paasonen, 2016), despite my professional experience.As a researcher, I'm aware that authenticity is fraught within the influencer industry: it's contingent on a number of individual negotiations among influencers and followers (van Driel & Dumatrica, 2020), and can be manufactured through affective labor (Abidin, 2016;Raun, 2018).But working through my event memories reminds me that this intellectual web of research on authenticity prioritizes the objective and academic over the affective-that gut feeling that something is genuine, even as you know it is probably not genuine at all.Abidin (2016, p. 12) finds that followers of family influencers often do not question the unpaid labor of the children featured in this content; rather, their performances of "raw, unfiltered" family dynamics are so convincing that followers seem to believe (to trust?) that they are "'family' before 'Influencers.'"Perhaps following an influencer is not about knowing whether something is real or not, but rather about feeling like you want something to be real.Much of this research on influencers and authenticity also (re)presents binary relationships with influencer content (authentic/inauthentic, trust/distrust, like/dislike) that oversimplify (and yes, overthink) the relationship between influencers and followers, as well as the experience of following itself.Following an influencer might mean choosing to trust their opinion on certain products but not their motivations, or it might mean trusting one video but not another; it might mean disliking the creator or content, but feeling drawn to them and following them anyway.

Conclusions
Following can seem transient and even arbitrary, but it's also important to acknowledge that who and what we follow is shaped by our online experiences.Since this project began, Dobrik has been accused of enabling and promoting sexual assault via his videos, and has lost some sponsorships as a result; the McBroom family has been accused of hiding their financial woes from followers, while Paul continues to troll his boxing opponents across various media formats and promote dubious business ventures.News articles about these scandals and each YouTuber's responses have been popping up in my Facebook feed as recommended content; while I have been following these YouTubers around, my online activity now means that they are now following me around.These algorithmically controlled social networks use data from my online activity to create a profile of my personality, needs, and interests, which they use to recommend content to me as a follower (Arthurs et al., 2018;Bucher, 2018;Saurwein & Spencer-Smith, 2021).As my algorithmic affect loop slowly encapsulated other social media platforms, it changed how and what I related to, what I knew about each YouTuber, and what I trusted.By tracing my research activity and feeding me further content about my research subjects to follow, these social networks enclosed me in an algorithmic feedback loop (Papacharissi, 2015), making it impossible to keep my personal online activity bracketed from that of my research.At the same time, my situated knowledge (Haraway, 1988) of following also gave me the tools to more adeptly follow following around, to more closely examine these feelings of relation, trust, and knowing.I therefore argue that in an era driven by algorithmic logics (Benjamin, 2019), this relationship has shifted from that of a binary, of researcher/follower, to a continuum: the researcher-follower.
In this paper, I attempted to unpack my experience of collecting and analyzing this data to reflect not only on my subjectivity (being a researcher, being a follower, how they can become entangled) but also on the larger theory behind how "YouTubers" or influencers work: how they form connections (Abidin, 2015(Abidin, , 2016)), why they are easy to connect to (Horak, 2014;Raun, 2018b), why it might be just as satisfying to not connect (Click, 2019;Gray, 2003Gray, , 2005)).Using Sinclair's (2020) event memories, I was able to see how I was making connections with these YouTubers and with my prior research, and how this follower/researcher tension led to moments of pleasure, recognition, frustration, and contemplation.Unpacking and tracing my own experiences of becoming a follower made me more aware of how (and what) I am aware (of) when conducting research.Following YouTubers such as Dobrik, Paul, and the McBroom family feels like (not) relating with them, cultivating knowledge about them, and (dis)trusting them.These YouTubers produce followers in part by making space for these feelings, for making it seem possible to relate, know, and trust them.This is part of what makes social media, and influencers such as these YouTube content creators, so fascinating; it's also what makes them terrifying.Much has been written about the spectre of social media influencing, which is turning people into passive consumers of online content but also commodities themselves (Carah, 2017;Moulard et al., 2015;Paasonen, 2016).However, following following around reminds me that it is still a decisively active process.Following can feel involuntary, but it consistently comes down to making a choice: to (not) relate, to (question what you) know, to (not) trust.Each of these actions shapes your orientation to these influencers and their content (Ahmed, 2010).
Mapping the phenomenon of being a follower also helped me think about what it means to be a social media researcher investigating algorithmic media (Arthurs et al., 2018;Lupton & Watson, 2021).This paper joins a growing body of scholarship exploring algorithmic media methodologies (Are, 2021;Brown, 2019;Lupton & Watson, 2021).Approaching this content with a phenomenological lens enables closer consideration of following as a singular and productive experience (Van Manen, 2014).I posit Sinclair's (2020) method of mapping event memories as a useful way to explore the phenomenon of algorithmic influence because it makes the researcher's own influence(s) more visible.Like algorithms, researchers mediate content based on internal biases, predictive impulses, and collaborative filtering (Airoldi et al., 2016;Benjamin, 2019).Virtual and digital technologies have expanded categories of data in qualitative research; however, regardless of its origins or typology, data is produced, selected, and analyzed to reflect the theoretical or methodological contexts of researchers (Flick, 2018;St. Pierre, 2011).Analyzing my event memories brings my subjectivity to the fore, (re)emphasizing that data are not inert, raw, or passive objects, but rather are dynamic and produced through entanglement with the researcher (St. Pierre et al., 2016;Torrance, 2019), especially in algorithmically mediated spaces (Lupton & Watson, 2021).Future research must continue to grapple with this methodological and epistemological quandary.