The intimacy triple bind: Structural inequalities and relational labour in the influencer industry

The careers of social media content creators, or influencers, live or die by their ability to cultivate and maintain an invested audience-community. To this end, they are encouraged to practise what has been framed as ‘emotional labour’ (Hochschild, 2002 [1983]) and ‘relational labour’ (Baym, 2018), commodifying their personalities, lives and tastes in order to build ‘authentic’ self-brands and intimacy with audiences. Drawing on an ethnographic study of the London influencer industry (2017–2023), this article examines emotional/relational labour through an intersectional feminist lens, foregrounding the ways in which structural inequalities shape relationships between creators and their audiences. The tolls of managing audience relationships are higher for marginalised creators – especially those making stigmatised and less brandable content genres – who find themselves on an uneven playing field in the challenges they face as well as the coping strategies at their disposal. These creators are in an intimacy triple bind, already at higher risk of trolling and harassment, yet under increased pressure to perform relational labour, adversely opening them up to further harms in the form of weaponised intimacy. This article explores four key tactics that creators employ in response to such conditions, as they navigate relational labour and boundaries with audiences: (1) leaning into making rather than being content; (2) (dis)engaging with anti-fans through silence; (3) retreating into private community spaces, away from the exposure of public platforms; and, in parallel, (4) turning off public comments. The adverse experiences of marginalised creators who speak about their identities and experiences online raise serious concerns about the viability of content creation as a career for these groups, as well as the lack of accountability and responsibility that platforms show towards the creators who generate profit for them.

Sometimes it seems like you're just there to be a Sims family playing house.The journey of your channel isn't about growing your creative style.It's about playing the game of life in public.You want some hit content?Get married, have kids, get a dog, do life.Do it all, do it right now as hard and fast as you can until your feet set on fire and we all laugh spectacularly when it burns to the ground!Howell (2022) In a 2022 video titled 'Why I Quit YouTube', OG British YouTuber Dan Howell reflects on the existential clash between 'making content and being content', identifying the latter as an essential ingredient for success in the influencer industry.With his characteristic deadpan humour, he gets right to the heart of what it means to work as a social media content creator: the commodification of the self.Dan is no stranger to such pressures, having been subjected to a decade of intense scrutiny from viewers about his sexuality and the nature of his relationship with his comedy partner Phil Lester since their rise to YouTube stardom beginning in 2009.He expresses deep ambivalence about this dynamic, recognising the potential success that comes with sharing aspects of his personal life, but also the pitfalls.In a previous video titled 'Basically I'm Gay' (Howell, 2019), he discussed the fraught nature of maintaining privacy as a social media star and the immense pressure he felt as someone with millions of followers to be a representative for the queer community, as well as his fear of the negative repercussions that inevitably come with being openly gay online.As he put it, 'I'm somebody who wants to keep the details of my personal life private. . .I want to do certain things without an audience' (Howell, 2019).Though a particularly high-profile example, Dan's reticence is not unique.On social media platforms where intimacy sells, every content creator grapples with how to present themself, how to relate to their audience, and how much to share of their life.
While it is well established in the literature that those who are variously termed content creators, influencers and microcelebrities 1 are required to practice 'emotional labour' (Hochschild, 2002(Hochschild, [1983]]) and 'relational labour' (Baym, 2018), there is a lack of attention paid to the unequal distribution of the tolls that managing these audience relationships can take.Situated within the broader context of the structural inequalities that mark labour in the influencer industry, this article seeks to build a dialogue between two seemingly disparate bodies of research: content creators' emotional/relational labour, on the one hand, and the proliferation of hate and harassment towards marginalised groups in digital cultures, on the other.I suggest that the imperative of emotional/relational labour and the flourishing of 'toxic technocultures' (Massanari, 2017) can -and indeed should -be thought together productively in the context of the influencer industry, as both concern the ways in which affect, pleasurable and painful, circulates between content creators and their audiences.
It is necessary to contextualise the argument presented in this article within the compounding structural inequalities that characterise the influencer industry.The advent of social media has provided a space for marginalised cultural producers as well as radical left-wing and feminist politics to flourish, and yet social media platforms are 'designed to render some content more visible than others, and the logic of this asymmetry is based on profitability', a system that rewards 'brand-safe' and depoliticised content and identities (Glatt and Banet-Weiser, 2021: 54; see also, Banet-Weiser and Glatt, 2023;Duffy and Meisner, 2022;Glatt, 2022b).There is a complex interplay between identity, politics and commerciality in the industry, with new forms of exclusions emerging due to the close alignment between the interests of advertisers and the governance of platforms, and the ways that this alignment unfolds in algorithmically structured environments.In platform contexts where creators are radically untethered from the support systems of traditional employment that protect against precarious working conditions (Glatt, 2022b), those making stigmatised and less brandable content genres, and especially content creators from historically marginalised groups, face barriers to earning a living and achieving visibility in the influencer industry, which map onto well-worn inequalities of race, class, gender, sexuality and ability.These creators face systemic exclusion from, and underpayment in, brand campaigns (Bishop, 2018a;Edwards, 2022;MSL, 2021), opaque and unpredictable age restriction and demonetisation on platforms (Glatt and Banet-Weiser, 2021) and heightened levels of hate and harassment (Duffy et al., 2022).Furthermore, feminist influencer scholars have explored how historically marginalised groups and non-normative expressions are punished by algorithmic systems in the creator economy (Are, 2022;Bishop, 2021a;Duffy and Meisner, 2022;Glatt, 2022a;Noble, 2018;Rauchberg, 2022).Unable to rely on income from brand campaigns, or platform ad revenue and creator fund schemes that depend on visibility and metric success, these creators often turn to their audiences for financial support via crowdfunding platforms like Patreon, and tipping apps such as CashApp, Buymeacoffee and Venmo.It is within this context of the uneven distribution of material and cultural resources in the influencer industry that my analysis of emotional/relational labour is situated.
The article is divided into four sections.I begin by introducing the literature on emotional/relational labour and related concepts as they pertain to working conditions in the influencer industry and the broader Cultural and Creative Industries (CCI), followed by literature relating to hate, harassment and marginalisation online.Proceeding to the analysis, I explore four key practices that marginalised creators employ in their attempts to manage relational boundaries with both friendly and hostile audiences: (1) leaning into making rather than being content; (2) (dis)engagement with anti-fans through silence; (3) retreating into private community spaces, away from the exposure of public platforms; and, in parallel, (4) turning off public comments.In the discussion, I bring these threads together, arguing that the relational tactics available to creators and their effectiveness need to be understood within the context of structural inequalities in the influencer industry.I present my conceptualisation of the intimacy triple bind: creators from historically marginalised groups -and especially those in stigmatised content verticals -are already at higher risk of trolling and harassment, yet under increased pressure to perform relational labour to sustain their income, which adversely opens them up to further harms.Thus, this article dovetails with recent scholarship that strives to understand the relationship between the requisite career visibility and the resultant public scrutiny, hate and harassment that is par for the course for content creators, 'all of which are exacerbated for women, communities of color, and the LGBTQIA community' (Duffy, Miltner, andWahlstedt, 2022: 1661; see also, Duffy and Hund, 2019;Duffy and Meisner, 2022).
As Baym (2018) argues, online workplace environments are populated by people you want to reach, but also 'those who hate you and those who love you too much'; and therefore creators need to present a self-appropriate to both home and work, or else find a way to keep identities distinct across contexts (p. 192).What this looks like for different creators varies greatly, and it is not my intention in this article to flatten out the range of experiences of creators who face diverse challenges, or to suggest that there is a universal model for capturing their approaches to emotional/relational labour.Instead, I am interested in exploring the patterned ways in which creators' identities and content genres shape their experiences with audiences, attending to questions of intersectional inequalities, with a view to understanding how the influencer industry provides both openings and foreclosures for specific kinds of cultural and political participation.
This research is ethnographically grounded.The data presented arose from a longterm study interrogating precarity and inequality for London-based content creators working in the influencer industry (2017)(2018)(2019)(2020)(2021)(2022)(2023), with the aim to address the platformisation of creative labour more broadly.This involved engaging with hundreds of creators through a range of complimentary methods: (1) offline participant observation at 10 key industry events in the UK and USA (2018-2020), as well as formal and informal content creator meet-ups and events; (2) online participant observation of creator culture across a wide range of social media platforms (YouTube, TikTok, Twitch, Instagram, Facebook, Twitter, Substack, personal blogs); (3) in-depth semi-structured interviews with 30 London-based content creators; and (4) autoethnographic research in the form of becoming a YouTube creator myself, with the aim of gaining firsthand experiential insights into the nature of content creator labour.The research encompassed fieldwork with a wide spectrum of entrepreneurial creators, from members of the London Small YouTubers, a community organisation for small creators (<20,000 subscribers) carrying out seemingly endless free 'aspirational labour' (Duffy, 2017: x), to elite A-list influencers in various green rooms and highly secured hotel bars at major industry events.Interviewees represented diverse identity categories (in terms of gender, race, sexuality, class and ability), worked across a wide variety of prominent and niche genres, and encompassed both full-time and aspiring creators, ranging from a single solitary subscriber (myself) to 2.2 million.This article draws on a selection of accounts from creators that pertain to the topic at hand, namely their experiences of, and approaches to, emotional/relational labour with their audiences, and how these illuminate dynamics relating to identity and marginalisation in the influencer industry.

The public game of life: intimacy and labour in the influencer industry
Not confronted with the usual gatekeepers of legacy CCI determining their suitability for opportunities and setting the disciplinary boundaries of their labour -line managers, commissioning editors, executive producers and directors -the careers of content creators in the influencer industry instead live or die by their ability to directly cultivate and maintain an invested 'audience community' (Baym, 2000).To this end, as many have noted, they are encouraged to commodify their personalities, lives and tastes, building 'authentic' self-brands, appealing on-screen personas, and intimacy with audiences (see, for example, Abidin, 2015;Baym, 2018;Bishop, 2018a;Cunningham and Craig, 2017;Duffy, 2016;Duguay, 2019;Raun, 2018).But the mobilisation of intimacy in creative work is not a new phenomenon.As Gill and Pratt (2008) summarise, several terms have been developed that speak to the relationship between broader transformations in advanced capitalism and the subjectivities of cultural workers, including 'creative labour, network labour, cognitive labour, affective labour and immaterial labour' (p.3).One aspect of this research concerns how cultural workers 'are required to deal with the particular emotions generated in the course of their jobs, both their own and those of others', shaped by the specific structural conditions of the CCI (Hesmondhalgh and Baker, 2008: 103).In this respect, many researchers have drawn on Arlie Hochschild's (2002Hochschild's ( [1983]]) influential concept of 'emotional labour' to describe activity whereby 'impersonal relations are to be seen as if they were personal [and] relations based on getting and giving money are to be seen as if they were relations free of money' (p.106; for example, Baym, 2018;Bishop, 2018a;Casey and Littler, 2022;Gill and Pratt, 2008;Grindstaff, 2002;Hesmondhalgh and Baker, 2008).
While this broader literature on creative labour is foundational for thinking about the work of social media content creators, it lacks an important element: that they are performers in an explicitly audience-facing industry, more akin to the celebrity status of actors and musicians than to the creatives working behind the scenes in industries such as television, fashion, and radio, or the service workers that Hochschild describes.In her book Playing to the Crowd (2018), Nancy Baym argues that the many terms used to modify contemporary labour are not adequate for understanding the changing relationships between musicians and their fans amid the rise of digital communication platforms, offering 'relational labour' as an alternative conceptualisation, which she defines as the 'ongoing, interactive, affective, material, and cognitive work of communicating with people over time to create structures that can support continued work ' (p. 19).More than the performance and creation of feeling, relational labour is about building and maintaining enduring relationships with audiences in a one-to-many setting, where performers 'must simultaneously manage the relational demands of each person who reaches them and play to the crowd as a whole, with all of the diverse audiences of allies, antagonists, strangers, and others it contains' (Baym, 2018: 20).Thus, she argues, any analysis of relational labour must also be understood in terms of the 'relational boundaries' that these creatives put in place, as they attempt to strike a balance between closeness and distance, and between work and life.
The influencer industry is fast approaching its 20th anniversary as a highly lucrative and established player in the global media marketplace, but its content, cultural norms and genres continue to be strongly shaped by its beginnings in amateur production, including the expected intimacy between video producers and consumers.Recognising YouTube's cultural roots, Raun (2018) argues that 'intimacy as genre presupposes [its] history. . .as an amateur-driven platform', with audiences expecting 'intimacy in content as well as style' through access to creators' everyday lives and private domestic spaces (p.105).As Abidin (2015) argues in her seminal piece on the 'perceived interconnectedness' between influencers and their audiences, 'the allure of influencers is premised on the ways they engage with their followers to give the impression of exclusive, intimate exchange'.Marwick (2015) describes the ways in which influencers utilise 'strategic intimacy' by presenting 'personas that appear to be less controlled than those of highly regulated, highly consumer brand oriented film and television stars ' (p. 346).As she puts it, 'while mainstream celebrities are expected to protect their privacy, micro-celebrities cannot or they'll lose this attention' (Marwick, 2013: 143).In her research on queer women's selfrepresentation on Instagram and Vine, Duguay (2019) found that they performed 'intimate affective labour' to tap into their audiences' assumed desire to see people like them expressing their sexuality within everyday contexts, an assumption that was corroborated through the high audience engagement with intimate content (p. 6).As Jerslev (2016) puts it, for content creators, work is characterised by 'continuous and multiple uploads of performances of a private self; it is about access, immediacy, and instantaneity' (p.5238).
While there are myriad personal and social reasons why content creators build intimate relationships with their audiences, there is no doubt that one motivation is economic.As Duffy (2016) highlights writing about feminised bloggers, 'aspirants recognize the instrumental value of their affective relations as they try to increase their followers and likes; improve rankings; and rethink approaches to content based upon feedback provided by their readers' (p.449).After all, the influencer industry is built on an ecosystem of advertisers, social media marketers and algorithmic systems, all of which are structured by the 'quantification imperative . . .wherein bigger typically translates into better' (Duffy, 2016: 449).In line with this interpretation, Bishop (2018b) argues that the anxiety disorder confessional vlog genre on YouTube can be understood as a tactic for increasing algorithmic visibility on a platform that rewards authenticity and eschews excessive commerciality (p.96).
Building on these works and Baym's conceptualisation, this article examines the relational labour of social media content creators through an intersectional feminist lens, focussing on how structural inequalities shape their relationships with their audiences, both friendly and hostile.While all creators must perform relational labour and manage relational boundaries as they attempt to strike a balance between closeness and distance with audiences, I argue that those from historically marginalised groups, and especially those whose content relates to identity politics, find themselves on an uneven playing field in the challenges they face as well as the coping strategies at their disposal.It is with this in mind that I turn to literature concerning harmful creator-audience interactions, such as trolling and harassment, framed by the ubiquity of racism, misogyny and transphobia online.

Toxic technocultures, misogynoir, and networked misogyny: the hate and harassment of marginalised groups in digital spaces
In her book How to stay safe online (2022), Seyi Akiwowo draws on a sad, but sadly unshocking, set of statistics: globally, women are 27 times more likely than men to be harassed online, Black women are 84 percent more likely to be harassed than white women, there has been a 71 percent rise in online disability abuse, and 78 percent of LGBTQ+ people have experienced hate speech online (Akiwowo, 2022).Moya Bailey (2016) coined the term 'misogynoir' in 2008 to describe the 'amalgamation of anti-Black racism and misogyny . . .that targets Black trans and cis women', particularly in visual and digital culture (p. 2).More broadly, Massanari analysed the flourishing of what she calls 'toxic technocultures' online, which are networked publics that 'demonstrate retrograde ideas of gender, sexual identity, sexuality and race and push against issues of diversity, multiculturalism, and progressivism' through the 'implicit or explicit harassment of others' (Massanari, 2017: 333).
Scholars have interrogated several dimensions of these phenomena.Massanari (2017) utilises actor-network theory to address the ways in which Reddit's design, algorithms and platform policies coalesce to provide fertile grounds for anti-feminist and misogynist activism and the 'ongoing backlash against women and their use of technology and participation in public life' (p.330).Taking a similarly ecological approach, Matamoros-Fernández (2017) explores how the entanglements of the national specificity of racism and the medium specificity of platforms constitute 'platformed racism', derived from the design, technical affordances, business models, policies and cultures of social media platforms.Understanding platforms and users to be mutually shaped, she highlights the dual meanings of platformed racism: first, platforms as tools for amplifying and manufacturing racist discourses, both through users' appropriations of their affordances and through their design and algorithmic shaping of sociability.And second, a mode of governance that is particularly harmful for some communities, characterised by vague policies, insufficient moderation and the arbitrary enforcement of rules.
While scholars have addressed various contributing factors to the 'especially virulent strain of violence and hostility' towards marginalised groups in online environments, Banet-Weiser and Miltner (2016) argue that accounts often fail to acknowledge the deep entrenchment and naturalisation of misogyny and racism as structuring logics in Western culture (p. 171).Focussing on specific components of 'networked misogyny', they contend, can serve to distract from the bigger cultural and normative fight at hand.This article is situated within the burgeoning field of Intersectional Critical Race Technology Studies, which provides a framework to interrogate the Internet as 'a system that reflects, and a site that structures, power and values' (Noble and Tynes, 2016: 2).In line with Noble and Tynes, I draw upon Patricia Hill Collins' (2015) definition of intersectionality to encompass 'the critical insight that race, class, gender, sexuality, ethnicity, nation, ability and age operate not as unitary, mutually exclusive entities, but as reciprocally constructing phenomena that in turn shape complex social inequalities' (p.32).I am also inspired here by the critical work of Angela Davis (1981), for her explicit disavowal of capitalism and the ways in which class relations intersect with other forms of marginalisation to exacerbate inequalities.It is with all this in mind that I proceed with my findings and analysis, understanding content creators' practices of relational labour and boundary-setting as shaped by platform cultures and affordances, but ultimately situated within, and inseparable from, broader structural inequalities in the influencer industry and society.

Relational labour and boundaries: an intersectional analysis of content creator practices
In the analysis that follows I unravel some of the issues that marginalised creators face and the tactics that they employ as they attempt to manage relational labour and boundaries with their audiences, both friendly and hostile.

'I have long resented being the product': making content versus being content
During this research, many creators expressed to me that the cultural imperative towards intimacy made them variously feel exposed, exploitative, and answerable to entitled audiences who felt that they deserved to have a say in their life choices.For some, the answer to this dilemma was to try to decenter themselves -their everyday lives, beliefs and relationships -and to lean more into making rather than being content.However, creators' motivations for and ability to enact this type of relational boundary-setting is shaped by their identity and content genre.In this section, I draw on the experiences of two English creators, Lucy Moon and Simon Clark.Both present as white, middle class and ablebodied, but Simon is a man working in the high CPM 2 topic-driven vertical of science EduTube, whereas Lucy is a woman creating lifestyle and fashion content.
Simon is a science creator whose YouTube channel gained popularity through vlogging about his life as a PhD student in atmospheric physics at Exeter University.When I interviewed Simon in October 2018 he was at a fork in the road, having just completed his PhD and trying to figure out his next career steps.He was acutely aware that his audience was drawn to his channel more for his personality, life and opinions than to learn about science, but he felt deeply uncomfortable about commodifying himself in this way.So, he had decided to actively move away from personality-driven content -even though, as he put it, he would make 'so much more money' from it -and towards creating a science education show that he could 'feel proud of', and was happier to monetise through advertising and merchandise.The month after our interview, Simon released a blog post that put it this way: I have long resented being the product, my face being the reason that people came to my channel rather than how I talked about things.Introducing a bit more distance between me and my ability to earn a living is definitely going to be beneficial for my mental health.(Clark, 2018) In a video released the following year titled A letter to myself ten years ago, Simon describes how as a vlogger his 'lived existence on Earth was a commodity.Something to be bottled, refined, and sold' (Clark, 2019).Because of this, he felt he had to constantly be 'market ready', which resulted in 'the strain to pretend that everything is fine at all times.And then the extra strain or fear of being found out, that you are selling a product that is not what it claims to be' (Clark, 2019).
Similarly, lifestyle and fashion creator Lucy told me during our interview in November 2018 that she had decided to take a step back from personality-driven content, though for her it was a direct response to unwanted audience interactions and critique.As she told me, April last year was probably the most successful that I've been, every video was getting like 150,000 views.I was doing really well, I was on trending all the time, and I was deeply unhappy because of all the attention I was getting.When you get positive attention and people fascinated with your life, you also get a lot of negative and a lot of uninvited commentary . . .It's really difficult to deal with because on an internet platform like YouTube, you can't avoid it.You are being fed it constantly: it's being tweeted at you, it's being commented on your videos . . .I can't deal with receiving it every day.(Lucy Moon interview, November 2018) Lucy had therefore decided to pivot her content towards fashion and beauty and away from talking about 'really controversial and very deeply personal topics' such as her issues with food, alcohol and sex.Although, as she said, female beauty and fashion creators actually get 'a lot more stick' and are 'scrutinised on a whole other level', she felt better able to cope with it because whereas previously audiences were scrutinising her 'whole life and elements of [her] personality', now the subject matter felt more removed.A further benefit of this pivot was the intentional weeding out of male viewers: I have pretty much 90% women watching me, which is perfect.I don't want men because I felt really unsafe when I had men following me and still now I get creepy comments, but they feel more distant . . .also now I don't have like 4chans where they're trying to find pictures of me in bikinis, thank God. (Lucy Moon interview, November 2018) Reflecting on Lucy and Simon's different experiences and approaches to making versus being content sheds light on the 'gendered authenticity policing' (Duffy et al., 2022) that female creators face in the influencer industry.Simon's decision to pivot away from personality-driven content was not based on fear of audience critique or harassment because of his gender, race, sexuality, ability or class, but rather to alleviate the internalised pressure to pretend that 'everything is fine' when it's not.For Lucy, on the other hand, her decision was a direct response to 'negative', 'unwanted' and 'creepy' comments from audience members, an example of the heightened scrutiny and that female creators are subjected to (Duffy and Hund, 2019).
Furthermore, despite the loss of income incurred by this pivot, in the uncontroversial and highly brandable vertical of science education -an overwhelmingly male genre both in terms of creators and audience members -Simon possessed the immense privilege in being able to decide between either making or being content without the fear of demonetisation and loss of sponsorship opportunities.This is not a viable option for creators whose content is less able to attract commercial sponsorship or who are subject to algorithmic discrimination.For example, Are (2022) argues that pole dancing is both a form of expression and a source of income, but that Instagram's algorithmic policing of women's nudity and sexuality through the shadowbanning of pole dancers reveals their ideological belief that it is 'risky, borderline and worth hiding in a way that . . .men's bodies and actions aren't' (p.15).Earning money on social media platforms is linked to content creators presenting sanitised and (hetero)normative versions of themselves and of the world, chiming with Hesmondhalgh's (2019Hesmondhalgh's ( [2002]]) argument that 'forms of communication that come to rely on advertising as their main source of income tend to become beholden to their advertisers' (p.281).

(Dis)engagement with anti-fans: harassment and the fear of weaponised intimacy
With 640k subscribers to their YouTube channel of 10+ years, Ash Hardell is one of the most famous and highly visible trans nonbinary content creators in the Western context and an example of a creator who has a very strong audience-community, having cultivated intimacy over the years via candid self-disclosure about their sexuality, relationships and experiences of transitioning, among other personal topics.In a video essay titled 'Trauma.Transphobia.And the Internet (why I left for 2.5 years)', they talk about the extensive harassment they received from social media audiences, as well as mainstream media, after their coming out video went viral in 2018, exposing their channel to a much broader audience than the predominantly LGBTQ+ viewers they had up until that point (Hardell, 2022).Ash describes how as their channel grew, YouTube's algorithms recommended their videos more widely to transphobic creators and their followers, drawing hostile audiences to their channel and creating an incredibly toxic environment, supporting Duffy et al.'s (2022) argument that far from protecting users from hateful expressions, platforms' algorithmic recommendation systems often inadvertently promote and reward this kind of content (p.1663).
The self-disclosure that Ash intended for their own audience-community thus became ammunition for anti-fans who were able to tap into Ash's deepest insecurities to a point where they developed post-traumatic stress disorder (PTSD) symptoms, and worse of all, as they put it, 'was gaslit into viewing myself as something dangerous . . .I was confusing and harming thousands of people who watched my videos, I was convincing cis people that they were trans and ruining lives'.These experiences left Ash with lingering trauma and shame, wary to share their personal experiences online in the future: I'm worried that by opening up and sharing my story and confessing to some of my personal self-doubts and struggles in the process, folks may try to weaponise those disclosures against me later.(Hardell, 2022) At its most extreme, weaponised intimacy can come in the form of doxxing, or the exposure of private personal information online, which can lead to risks for the recipient, including physical harm.In 2018 Ash was the victim of a doxxing attack, in which it was publicly shared which airline they were working for as a flight attendant, in a video which amassed more than 500k views in the following months.Ash spoke to an entertainment lawyer, who explained that in situations of doxxing, most creators do nothing.Pursuing legal action is lengthy and expensive, and more importantly, any attempt to contact the doxxer or respond publicly almost always results in more controversy and traffic to the original post or issue: 'So basically the more you try to address or handle a problem, the harder it becomes to escape it.What a broken system'.(Hardell, 2019).Thus, the best way to defend against online attackers becomes silence: creators are pressured not to respond to harassers and doxxers or talk about anything personal publicly online that could be weaponised.For Ash, the doxxing and barrage of transphobic messages that they received were so intense that they were driven away from YouTube altogether for more than two years, before building up the resolve to return and address what had happened out of a sense of duty to challenge the 'far-reaching and diverse hatred towards trans people online' and to get 'the closure that comes with having control over [one's] own narrative' (Hardell, 2022).
On a panel about influencer mental health at Summer in the City 2019, the UK's largest community-oriented online video conference, one Black creator said that 'trolling and hate just comes with the territory and you have to develop a thick skin to work as a social media creator'.I heard this trope many times throughout my fieldwork and interviews, particularly from marginalised creators and those whose content deals with identity politics, highlighting the individualisation of risk and harm as a structural norm in the influencer industry.
'Those are my people': retreating into private community spaces and dealing with obsessive fans I met Ahsante Bean (AKA Ahsante the Artist) on my first fieldwork trip to VidCon US in 2018.She is a Black creator who has been making YouTube videos since 2013 about personal growth, productivity, intersectionality, identity and social justice, with the goal to share stories that inspire her audience-community 'to move consciously and creatively through life' (Bean, 2022).She is also the former Associate Director of Programming at PBS Digital Studios, 3 where among other things, she created Say It Loud, an educational series on YouTube celebrating Black history and culture.Ahsante told me in our interview in August 2022 that as her audience had grown, the quality of conversation in her YouTube comments section had diminished, with an influx of hostile audience members on some of her more controversial videos which were 'disrupting the sense of community in the comments' and 'mostly just grandstanding about their own beliefs and trying to use my platform to promote whatever their agenda was'.
As a result, in July 2022 Ahsante posted a video on her channel titled 'I'm leaving . . .come with me', in which she explains her decision to pivot to a new YouTube channel focussed solely on video essays, in parallel with a migration of her audience interactions away from the YouTube comments section and towards her Patreon, which she justifies in this way: I want it to be focussed on having a consistent dialogue with folks who are in my corner, who know and love my work, and who want to support the quality of my creative career . . .Going forward [Patreon] is going to be the main place where I'm engaging and commenting and discussing my ideas with people who really care about them.The public internet can be a ruthless, hypercritical place.YouTube comment sections can be pretty dicey.(Bean, 2022) Many creators retreat into the private community crowdfunding space of Patreon in a bid to communicate with a smaller and more invested subsection of their audience, while avoiding negative and unproductive audience interactions on public platforms.Patreon has the additional benefit of creating an audience-centric revenue stream not directly tied to algorithmic visibility, which is particularly important for creators whose content/identity is not easily brandable in the influencer marketplace and who face regular algorithmic discrimination.For Ahsante, this meant she could pursue a 'quality over quantity' approach, against YouTube's imperatives to publish or perish and to put out 'attention grabbing' (read: clickbait) content.
Crowdfunding also aligned with Ahsante's personal investment in the values of public service media, in that a supportive subsection of her audience (her Patreon community) invested in her career, which made it financially viable for her to continue to create and publish free and publicly available educational content on YouTube without relying on brand sponsorships.As she put it, 'Whoever is paying you, that's who you're serving, that's the client . . .If brands are paying me then that's who I'm accountable to, that's who I have to satisfy and make happy, and I don't want to do that'.For Ahsante, the crowdfunding model enabled her to be accountable to and prioritise her audience without having to worry about selling herself and her audience's attention to brands.This decision was made easier by the fact that she had only had positive experiences with her patrons to date.This aligns with the experiences of participants on the Patreon panel at VidCon UK 2019, who explained that while Patreon is structured around the idea of audiences paying creators in exchange for special perks (such as merchandise and exclusive content), many viewers become patrons because they understand the precarity of content creation as a career and want to provide financial support to their favourite creators, as well to become part of their 'inner circle'.
For others, the relational labour of cultivating an intimately invested audience to whom they are directly accountable is extremely taxing.For example, one trans 'BreadTube' creator -a critical leftist video essay genre addressing such topics as philosophy, race, gender, capitalism and politics -told me in our interview in October 2019 that her decision to bring deeply personal topics into her videos had generated an 'overwhelmingly positive' response from her audience and helped to foster a 'wonderful community'.However, she said that the major downside of this intimacy was that it had given rise to not one, but 'a few' stalkers.She had never gone to the police over death threats, of which she had received many, but she did feel it necessary to go the police about a stalker who was 'obsessed' and thought they were 'in some kind of relationship'.We can understand obsessive fan behaviour like stalking as another form of weaponised intimacy, different but no less harmful than hate-fuelled iterations.This creator's approach mirrored Livingstone's (2008) findings in her study of teenagers' use of social networking sites, that they 'must and do disclose personal information in order to sustain intimacy, but they wish to be in control of how they manage this disclosure' (p.405).She used the fact that she created scripted content, which often blurred the line between fact and fiction, as a tactic for maintaining relational boundaries with her audience: she could always 'write stuff off as a joke'.For example, when she had used BDSM gear in a video, audience members speculated whether she already owned it or had bought it as a prop, but, as she told me gleefully, 'They'll never know'.She was thus able to use this ambivalence to diffuse some of the more intense expectations of intimacy with her audience.

'The public Internet can be a ruthless, hypercritical place': turning off comments
In a break with the norm, Ahsante had made the bold choice to turn off comments completely on her new YouTube video essay channel.While it's common knowledge that YouTube comments sections are a prime location for hate and harassment, especially for marginalised creators, it is extremely unusual for creators to turn off the comments, which highlights the pressures they are under to open themselves up to public scrutiny in the course of engaging in relational labour with their audiences.Ahsante, however, had made peace with this decision.She had big plans to grow her new channel to a significant size by creating video essays dealing with controversial topics such as race and intersectionality, and therefore expected harassment.She had dealt with this in the past when her videos about race had been pushed out by YouTube's algorithms to a wide audience, which had resulted in 'people sending their army over and the comments getting deluged' and had found that disabling the comments was the most effective and absolute means of shutting down the harassment of herself and her audience-community.Though comment moderation is not currently a requirement of content creators enforced by most social media platforms (with the noteworthy exception of Twitch), informed by her background in journalism, Ahsante felt an ethical duty towards her audience to provide 'a productive and safe space' in the comments section.This was only achievable through active moderation, which she was unable and unwilling to spend hours of her time performing or pay someone else to do.As she told me, 'It's the Wild West down there . . .A lot of publishers had a moment where everyone turned the comments section off because they realised it's just a garbage fire . . .but it's a different norm with influencers and YouTubers'.
Ahsante's decision to leave her videos publicly available on YouTube but to move her audience interactions behind the Patreon paywall and set the unusually firm boundary of switching off public YouTube comments can be seen as the logical conclusion of the approach of Duguay's (2019) participants, who considered switching to other platforms with 'more privacy features' because of the affectively draining process of constantly having to delete comments and block/report users (p. 6).

Discussion and conclusions: the intimacy triple bind for marginalised creators in the influencer industry
In the afterword to the twentieth-anniversary edition of The Managed Heart, Hochschild (2002Hochschild ( [1983]]) reflects on a growing third sector of social life, which she terms 'marketised private life', a merging of the public work and private family cultures that the original book was structured around.We can understand content creators as examples par excellence of marketised private life, practising what Casey and Littler (2022) term 'digital identity labour', 'in which the identity of the self is crafted, commoditized and monetized for profit' (p.496) -most commonly through relational labour, with creators incentivised to form intimate connections with their audiences -marking a profound collapse of the already murky boundary between work and play endemic to labour in the CCI (Gill and Pratt, 2008).
While relational labour and boundary-setting are fundamental to the work of all content creators, this research found that the tolls of managing audience relationships are higher for marginalised creators -especially those whose content deals with identity politics -who find themselves in what I call the intimacy triple bind: (1) these creators face complex systemic technological, sociocultural and commercial exclusions that impact their visibility and income-generating opportunities; (2) they are therefore under increased pressure to rely on audiences directly for financial support via crowdfunding platforms like Patreon, and tipping apps such as CashApp, Buymeacoffee and Venmo; (3) however, for these creators, who are already at high risk of hate, harassment and doxxing, the imperative to perform relational labour required for the audience-dependent income model comes with higher risks to their mental health and safety, in the form of weaponised intimacy from both hostile and enamoured audience members.We can see particularly visible examples of this dynamic in trans female BreadTube creators, who are doubly marginalised on the basis of both their identities and their content and who rely heavily on their audiences for financial support (Glatt and Banet-Weiser, 2021).
As Baym (2018) puts it, 'the new demands of intimacy can be too much: too commercial, too much time, too much interaction, too much expectation, too much vulnerability, too much risk ' (p. 178).A throughline between many of the creators discussed in this article is that the personal risk and harm that comes with exposure to very large audiences makes their job untenable.It is clear that despite the 'visibility mandate' (Duffy and Hund, 2019) that structures the influencer industry, many of the tactics that marginalised creators employ involve some sort of retreat: away from confrontation, away from virality, away from the public Internet, or away from audiences altogether.This chimes with the work of Chan and Gray (2020), who found that in the face of the exclusions that Black men experience when building visibility and fame as online gamers, many have 'shifted their energies away from seeking public attention and building massive platforms', instead embracing their microcelebrity status within a community of other Black streamers and audience members (p.359).Hochschild (2002Hochschild ( [1983]]) argued that a certain degree of disengagement is required to manage the psychological tolls of emotional labour and maintain the sustainability of these jobs: 'The only way to salvage a sense of self-esteem. . . is to define the job as "illusion making" and to remove the self from the job, to take it lightly, unseriously' (p.93).But the vulnerability of being a social media content creator, a job that requires the cultivation of intimate connections with audiences -of disclosing deeply personal aspects of their lives, identities, beliefs and experiences, which may then be weaponised against them -is next to impossible to take lightly.However thick one's skin may be, the harassment, death threats, doxxing and hate speech that are common experiences for many marginalised creators who speak about their identities and experiences online raises serious questions about the viability of content creation as a career for these groups, as well as the lack of accountability and responsibility that platforms show towards the creators who generate profit for them.

ORCID iD
Zoë Glatt https://orcid.org/0000-0003-1665-7309 The author(s) disclosed receipt of the following financial support for the research, authorship, and/ or publication of this article: This work was supported by the UK Economic and Social Research Council (ESRC) Doctoral Training Partnership Studentship.