The Nested Precarities of Creative Labor on Social Media

Abstract

While metrics have long played an important, albeit fraught, role in the media and cultural industries, quantified indices of online visibility—likes, favorites, subscribers, and shares—have been indelibly cast as routes to professional success and status in the digital creative economy. Against this backdrop, this study sought to examine how creative laborers’ pursuit of social media visibility impacts their processes and products. Drawing upon in-depth interviews with 30 aspiring and professional content creators on a range of social media platforms—Instagram, YouTube, TikTok, Pinterest, and Twitter—we contend that their experiences are not only shaped by the promise of visibility, but also by its precarity. As such, we present a framework for assessing the volatile nature of visibility in platformized creative labor, which includes unpredictability across three levels: (1) markets, (2) industries, and (3) platform features and algorithms. After mapping out this ecological model of the nested precarities of visibility, we conclude by addressing both continuities with—and departures from—the earlier modes of instability that characterized cultural production, with a focus on the guiding logic of platform capitalism.

Introduction

In early 2020, as the global pandemic began to wreak havoc on contemporary labor markets and economies, a number of high-profile publications began to prognosticate the fate of a relatively new category of cultural producer: social media creators. Rather hyperbolically, a feature in Vanity Fair headlined, “Is this the end of influencing as we knew it?” (Bryant, 2020). A senior writer at Wired meanwhile, forecasted the uncertain career prospects for professional Instagrammers given that “brand deals have dried up; sponsored posts have been delayed” (Pardes, 2020). Other analysts highlighted the temporality of the platforms themselves, and—in particular—TikTok, which was mired in controversy with a rumored sale and threats of political bans in both India and the United States. In an MIT Technology Review report on the platform’s shaky future, Abby Ohlheiser (2020) summarized, “Whether the app goes or stays, this moment is forcing creators . . . toward a realization that making, or even consuming, things on the Internet means depending on platforms that could change drastically in an instant” (para. 5). Then, in a nod toward the fickle algorithmic systems that structure creator-centric platforms, she added, “Small algorithmic changes by a platform can make or tank an entire career” (Ohlheiser, 2020).

Taken together, such reports underscore a fundamental reality of careers borne on social media: they are profoundly volatile, bound up with fluctuations in wider socio-economic, cultural, and political realms. Such instability, of course, well predates these platforms; that is, careers in the creative industries have long been marked by intermittent or piecemeal work, low and/or deferred wages, and the cautionary mantra that “you’re only as good as your last [gig]” (Blair, 2001; Gill, 2010; Hesmondhalgh & Baker, 2013; McRobbie, 2016). Indeed, more than 15 years ago, Neff et al. (2005) described how new media workers face patterned “uncertainty and lack of control”—forms of risk offset by what workers view as “the promise of one Big Job being right around the corner” (p. 319). But today’s platformized creative economy—powered by a neoliberal ethos of self-commodification, the recasting of independent employment as “entrepreneurship,” and an oversaturated supply of talented hopefuls—amplifies many of these less auspicious features (Glatt, forthcoming; see also, Bishop, 2019; Duffy & Sawey, 2021). In this article, we call attention to a particularly acute source of such unpredictability—namely the much-hyped ideal of social media visibility—by offering an ecological model to assess its impact on creative laborers and their work.

Notions of visibility carry hefty analytical weight in contemporary writings on platforms (e.g., Bucher, 2012; 2019; Gillespie, 2018; van Dijck & Poell, 2013) and platformization (e.g., Helmond, 2015; Nieborg & Poell, 2018). Bucher’s (2012) exposition of a Foucaldian “regime of visibility” governed by the logic of algorithmic systems is especially relevant to the present study. But while Bucher’s initial formulation focused on the experiences of social media users, the stakes of being rendered visible—or conversely, invisible—are intensified for platformized creative laborers given that their material gains are hitched to metric success (Glatt, forthcoming). Through a case study of those selling their wares on arts-and-crafts e-commerce sites like Etsy, Klawitter and Hargittai (2018) argue that, in the digital economy, these kinds of workers “are more incentivized than most in understanding how [algorithmically curated] feeds decide what to display to users given that their livelihood directly depends on such exposure” (p. 3504). Similar conclusions have been drawn from studies of creative workers across mainstream platforms, most notably Instagram (Arriagada & Ibanez, 2020; Cotter, 2019; O’Meara, 2019; Petre et al., 2019) and YouTube (Bishop, 2019, 2020; Cunningham & Craig, 2019; Caplan & Gillespie, 2020).

Despite the rich insights of these works, these studies tend to focus on creators’ experiences with a particular platform—a trend that has elided attention to the role that cross-platform branding logics (e.g., Scolere et al., 2018) play in the production of digitally mediated personae. ¹ Moreover, given the cultural industries’ decades-long reliance on metrics as proxies for quality and mechanisms for feedback, it seems necessary to develop a perspective on visibility that acknowledges the broader forces that have historically impacted creators in traditional cultural labor contexts. Accordingly, this study sought to develop a broad ecological understanding of how creators working on multiple social media platforms imagine the visibility of the platforms on which they create and circulate content, with particular attention to the impact on their processes and products. Data for this project included in-depth interviews with 30 aspiring and professional creative laborers on a range of social media platforms, including Instagram, YouTube, TikTok, Twitter, and Pinterest. These interviews were conducted across a span of more than 2 years; hence, they provide insight into creators’ experiences with visibility across key moments amid what scholars describe as the continuous stages of platform evolution (e.g., Arriagada & Ibanez, 2020; Poell et al., in press).

Our findings reveal that not only do the challenges of maintaining visibility extend to the digital economy but that, further, both the promises and precarity of visibility are amplified in many platformized labor contexts. We posit, moreover, that creative laborers experience unpredictability across three levels: (1) markets, including evolutions in audience tastes, advertiser demands, and sources of competition; (2) industries, including evolutions in the ecology of platforms on which creators produce and circulate content; and (3) platform features and algorithms, including evolutions in the platforms themselves, ranging from transparent changes (e.g., new features) to more opaque changes (e.g., changes to the algorithm). As we show, these levels are deeply interwoven, underscoring the need for an ecological model of precarities. After offering such a model to help illuminate the nested precarities of visibility, we conclude by addressing both continuities with and critical departures from the earlier modes of instability that characterized creative work. That is, while cultural producers have long reckoned with the unpredictability of market demands and audience tastes, today’s precarity is largely an upshot of platform capitalism (Srnicek, 2017), rendering creators beholden to Silicon Valley’s every whim and vagary.

Creative Labor in Context: Metrics, Measures, and Markets

As academic writings on Instagram influencers, YouTube creators, professional bloggers, and DIY musicians attest, social media content creation is a category of labor that—while novel—bears a marked resemblance to work in the legacy media and cultural industries (Cunningham & Craig, 2019; Glatt, forthcoming; Hesmondhalgh et al., 2019; Homant & Sender, 2019). Accordingly, it seems important to trace the roots of the “new modalities of visibility engendered by new media” (Bucher, 2012, p. 1164) to earlier systems of measurement that scaffolded traditional fields of cultural production. Accounts of the “digital reputation economy,” for instance, have identified antecedents in earlier systems of ranking and rating cultural products, such as music charts and best seller lists (Hearn, 2010, p. 428). Gillespie (2016), meanwhile, notes that media industry participants have long deployed “both subjective assertions and manufactured metrics about what’s popular, what’s about to be, and what should be” (p. 62; see also, Baym, 2013).

As such patterned references to popularity indicate, audience size has long served as a significant—albeit deeply fraught—barometer of cultural tastes. And, correspondingly, the increasingly specialized mechanisms of audience measurement that developed in the second half of the 20th century shaped both the content and format of media systems (Turow, 1997). More pointedly, the metric systems furnished by Nielsen, the Audit Bureau of Circulation, and companies of their ilk engendered what analysts have productively described as “industrially constructed images of the audience” (Ettema & Whitney, 1994; Turow, 1997). As the language of constructionism attests, the media and advertising industries’ conceptions of audiences were understood as cognitive shorthand; creators thus treated them as suspect, at best. For instance, Gitlin (1983) details prime-time TV executives’ deep ambivalence toward audience ratings. Amid the industry’s increasing mandates for “hard knowledge” through concept testing, social research, and Nielsen ratings, TV workers became, paradoxically, ever more attuned to the “limits of this knowledge” (pp. 31, 42–43). Less than a decade later, Ang (1991/2006) argued that the TV industry’s incessant quest to lure audiences was itself a symbol of the “unknowable” nature of audiences. Consequently, the television institution was “[haunted by] a constant sense of uncertainty” (p. 18).

These and other accounts of legacy media industries make clear that cultural producers’ charged relationship with audience measures—and wider systems of quantification—presage the rise of algorithms and Big Data. But the digital media economy’s tremendous provisions of data have undoubtedly impacted the creation, circulation, marketing, and monetization of cultural content (Poell et al., in press). As we show in the following section, logics of personalization (i.e., content created for “an audience of one”) and ubiquitous algorithmic culture are at the forefront of assessing this impact.

Visibility in Platformized Cultural Production

While contemporary newsrooms have been described as “traffic factories” (Petre, 2015), entertainment-based industries are increasingly reliant on automated systems of demand prediction and evaluation (e.g., Napoli, 2014). Such a robust reliance on quantification can be understood as both a cause and symptom of the so-called platformization of the cultural industries, defined by “the economic, governmental, and infrastructural extensions of digital platforms” into these industries along with interrelated shifts in their practices of labor, creativity, and democracy (Nieborg & Poell, 2018; Poell et al., in press). It is against this backdrop that scholars have examined how metric logics are shaping the experiences and output of platformized workers across the news (e.g., Christin, 2018; Petre, 2015), music (Baym, 2018; Hesmondhalgh et al., 2019; Jones, 2020; Morris, 2015), and television (Lotz, 2014) industries. Importantly, the command to pursue quantified indices of visibility is no longer the exclusive province of complex organizations like newsrooms and TV studios. Amid an individualized digital creative economy, metric evaluations are increasingly legible at the personal level, most notably through social media platforms. In 2013, Baym (2013) described the rapid uptake of social media by cultural producers as evidence of a “larger shift toward the markers of profitability . . . less tied to production and more tied to reputation” (n.p.; see also, Gandini, 2016; Hearn, 2010). She added, “[I]t has also become important to measure audiences in order to assess social values such as legitimacy, credibility, likeability and other kinds of status that can presumably (if mysteriously) be converted into economic capital” (italics added, Gandini, 2016; Hearn, 2010).

These quantifiable “assess[ments] of social values” are not just discernible but, rather, instrumental, in platformized creative labor. Here, current and aspiring cultural workers are roused to pursue quantifiable markers of visibility—from likes and favorites to shares, pins, and comments—in earnest (Bishop, 2019; Duguay, 2019; Duffy & Sawey, 2021; Petre et al., 2019; Scolere & Humphreys, 2016). A number of studies in this vein have focused on Instagram influencers and aspirants; scholars have thus described Instagrammers’ perceived obligation to engage in “visibility labor” (Abidin, 2018); to play “the visibility game” alongside algorithms and platforms owners (Cotter, 2019); to “optimize their visibility” (O’Meara, 2019); and to adhere to gendered prescriptions of the “visibility mandate” (Duffy & Hund, 2019). Studies of the YouTube community, meanwhile, detail creators’ efforts to evade “the threat of invisibility” (Bishop, 2019) while accommodating platform logics configured by “visibility, metrics, and hierarchy” (Glatt, forthcoming; see also, Homant & Sender, 2019).

Collectively, such accounts make clear that visibility is a requirement for career success in cultural industries amid platformization, and that, accordingly, metrics are the central axes on which power and resources are exchanged—in the form of sponsorships, brand deals, collaborations, and so forth. Many of these works, moreover, underscore the role of (perceived) algorithmic systems in configuring creators’ level of visibility—often in ways that are lopsided, if not outright biased (Caplan & Gillespie, 2020; Bishop, 2019). Algorithmic systems are dynamic by definition; as such, the nature and extent of their impact seems to change at whim, alongside other forms of the so-called platform evolution (Arriagada & Ibanez, 2020). Such changes often necessitate that content creators “adapt their brand subjectivities and practices across platforms and affordances” (Arriagada & Ibanez, 2020, p. 2). As such, it is critical to assess the impact of visibility across a broad timeframe and, we would add, across the span of creator-centric platforms, given the guiding logic of cross-platform branding (Cunningham & Craig, 2019; Glatt, forthcoming; Scolere et al., 2018).

Methods

This project involved in-depth interviews with 30 aspiring or professional platformized creative laborers who create and/or circulate content on Instagram, Twitter, YouTube, Pinterest, and TikTok. ² Interviews focused on workers’ experiences within and across particular platform contexts, including their pursuit of visibility and their understanding of the forces that aided or impeded this pursuit. To recruit participants across platforms, career levels, and content genres, we relied upon three concurrent strategies: (1) directly contacting people who positioned themselves as knowledgeable about social media visibility in creative work; (2) contacting those who publicized their experiences with creative labor on platforms; and (3) utilizing a snowball sampling approach to solicit additional potential participants at the completion of each interview. ³ While most of our study participants focused their time, energy, and labor on a single platform, the majority also found it necessary to maintain a brand presence across the wider social media ecology—a theme we return to in the findings section. Additional details on the interviewees—each of whom was given a pseudonym to protect their privacy—are provided in the Appendix.

We initially conducted two rounds of interviews (in 2019 and 2020); subsequently, we opted to include a secondary analysis of data collected in 2018 for a separate study (Duffy & Sawey, 2021) given the similar topics and orientations to visibility. Analyzing data across multiple years proved generative to the study, in part by providing insight into creators’ experiences across key moments in the evolution of platforms. Some of these moments emerged as especially salient to creators’ pursuit of visibility; among them were Instagram’s continuous tweaks to its algorithms and its decision to “hide” likes; political controversies that sullied the reputations of Twitter and Facebook; and the rapid uptick in TikTok use attributed to the COVID-19 pandemic.

Due to the geographically dispersed nature of our sample, interviews took place over Skype and/or the phone and were audio recorded with participants’ permission. Interviews followed a semi-structured interview protocol; most lasted between 30 min and 1 hr. At the completion of the interview, participants received a stipend (US$25) in exchange for their time and insight, and interviews were sent to a professional service for transcription. The study’s authors independently read all of the transcripts and met regularly to discuss themes that emerged inductively (Glaser & Strauss, 1967). Together, the collaborators developed the coding categories and applied focused codes to the dataset. Finally, the researchers discussed the focused codes and organized them into overarching themes, which are discussed below.

Findings: The Promise and Precarity of Visibility

Given the logic of visibility that propels the social media economy, it is perhaps not surprising that participants detailed inclinations to have their content “seen,” to “build an audience” and “get attention,” to craft “posts that get more traction,” and—to use an oft-repeated euphemism for metric success—“do well.” Crucially, such motivations were overwhelmingly cast in terms of professional advancement, rather than personal self-interest. For example, James distinguished career-minded TikTok creators like him from those social media users who “post something, and [when] they get a lot of likes, or a lot of engagement, it gives [them] an ego-boost, but I don’t think they really care about it.” To James, who generated humorous content tailored to the LGBTQ+ community, “car[ing]” about the number of likes meant being strategic about one’s content and the attention it draws.

In chronicling their pursuits of visibility, participants offered wildly different metric targets (i.e., “hit a thousand [likes] for every picture” or reach “5,000 followers before my birthday”). Some, like Instagrammer Naina, sought to reach follower- and engagement- thresholds that she expected would lure brand sponsors. She explained, “If you can get a 50 percent engagement rate . . . that’s probably so much better to a brand.” Others, meanwhile, discussed how achieving metric benchmarks would enable them to curry favor with platform companies, particularly in cases where the latter supplied creator-centric funding initiatives. Olys shared her belief that her fellow Instagrammers sought to reach 10,000 followers to take advantage of platform features, “Once that number hits 10k, whether your followers are organic or not, you can fully take advantage of having a business account.” To Olys, this was a key impetus for those who engaged in ostensibly surreptitious practices, such as “buying followers” or other efforts to game the system (Petre et al., 2019; Ziewitz, 2019). Several TikTok creators, meanwhile, reflected on the platform’s standards for admittance into its buzzy Creator Marketplace. ⁴ Of their future goals on the platform, drag performer Alex offered: “I would love to increase my followers, definitely. Because once I hit 100K, I can then join the Creator Marketplace which will help me get into what I really want to do which is, like, working with brands, promoting products, and being an influencer. . .”

Despite such concerted efforts to boost their visibility, creators expressed frustration when their content failed to register what they considered a successful impact on audiences. In recalling what drove her to TikTok, Kavianah said, “I just really couldn’t understand why [my numbers on Instagram weren’t increasing] . . . I feel like I’m not being seen.” Together, such accounts underscore a fundamental axiom of platformized creative labor, namely that it is structured by both the promise and precarity of visibility. Participants attributed such unpredictability, moreover, to a patterned set of external factors, from the whims of audiences to platforms’ notoriously inscrutable algorithmic systems. In what follows, we show how these factors were embedded in structures of precarity at three levels: (1) market precarity, or the volatile changes in audiences’ tastes and advertisers’ demands; (2) industry precarity, or uncertainty that stemmed from the hyper-competitive ecology of mainstream creator platforms; and (3) platform precarity, or the changes enacted by individual platform companies in the form of updates to their features and algorithms. Although we present these levels discretely, evolutions in markets, industries, and platforms are by no means disconnected. Rather, as we show in the conclusion, they constitute the nested precarities of visibility. ⁵

Market Precarity: Audiences, Trends, and Social Shifts

Interview participants spanning platforms and genres offered a similar perspective on the market for social media content: it was fickle, elusive, and, ultimately, difficult to predict what would “do well.” As James noted of the erratic nature of popularity on TikTok: “I think it’s just about what people are watching, what they want to see, and what they identify with.” He added, “As a creator, that’s the hard part: ‘What are people going to engage with? What’s gonna resonate with them?’” TikTok entertainer Cooper used the language of lost control to describe the uncertainty of content visibility on the platform:

It’s hard to predict [what will get attention]. You don’t know. So, I’m always like, “Once I post a video, it’s out of my hands. I have no control.” I’ll do everything I can when I’m creating a video to make sure that this is as entertaining as it possibly can be as the content that it is, but beyond that, I have no control.

As a buffer against such uncertainty, creators relied upon metric data from past content as well as analytics provided by the platforms. As aspiring Instagram influencer Eva shared, “I’m starting to learn more about my audience and what they want, and that works better for me too in terms of growing.”

Angela, similarly, noted how members of Instagram’s creator community tend to invest their energies in content most likely to “go viral”; she clarified, “stuff that is getting just a jillion likes.” Angela, who recruits influencers, suggested that certain types of content—such as images that repurpose popular movie quotes—are especially likely to ratchet up metrics, but she also added the following caveat: “that changes over time.” Renee, accordingly, noted how she shifted the focus of her Instagram account toward wellness content to tap into what she described as “the big yoga/meditation trend going around the world right now.”

Despite creators’ reliance on data on audiences and content “trends,” such knowledge was deemed ineffectual at especially transitory social moments. Indeed, the volatility of market trends was brought into sharp relief during our final round of interviews, as a number of creators detailed the presumed impact of the pandemic and Black Lives Matter protests. James, for instance, recalled how in response to the Black Lives Matter protests, “[TikTok was] pushing out videos from protests. They were pushing [content from] Black creators.” While James lauded the platform’s decision to “amplify [marginalized] voices,” he acknowledged this was likely a business decision: “Whatever is relevant, and trendy, in that moment, I think that they adjust the algorithm . . . to push that stuff out more.” As we discuss in a later section, such algorithmic “adjust[ments]” were a key source of unpredictability.

Other TikTok creators, meanwhile, noted how pandemic-related stay-at-home orders led to an unexpected boost in their metrics. Nickolas, who created content in the tech space, recalled the “influx” of creators on TikTok during the first months of quarantine, saying, “I’ve seen teachers, doctors, other educators, and other medical professionals on the platform.” Consequently, the culture of the platform underwent a marked evolution during this time:

It used to be an app just for jokes, like people would lip sync to songs, and it would primarily be a dancing app where everyone pretty much would post dancing videos. But I would say right now it’s sort of evolved to where you can find a wide variety of content on the app. And I guess more people posting educational content has really helped . . . it has definitely changed throughout the time that I’ve been on the app.

Despite the unprecedented surge in usage of the platform, Nickolas explained that many brands remained reluctant to invest in advertising. He explained, “some brands are still not considering TikTok as a viable platform to market on.” Nolan, who circulates tech and computing content, by contrast, offered a different perspective on advertisers’ assessment of the short-form video app:

[Brands] . . . just saw potential that the younger generation on here has . . . The creators influence them a lot, so they probably saw a lot of potential just on TikTok, just because it’s a new platform.

Such comments make clear that the precarities of markets—especially those that impact the visibility of creators—reflect wider external forces, including audiences, advertisers, and the socio-political environment for cultural goods.

Industry Precarity: Ecology of Platforms

At the same time that participants described the instability wrought by ebbing and flowing markets, they also noted the unpredictable nature of the platform ecology. Over a span of mere years, some creator-centric platforms like TikTok and Clubhouse launched or evolved while others (e.g., Vine) seemed to vanish overnight. During our second wave of interviews, Instagrammer Eva noted how the ever-changing nature of individual platforms led to reconfiguration of the wider social media ecology:

[Constant change is] just something you have to agree to when you start using these things. And accept things are always gonna change, social media changes all the time . . . So, you either just get really angry at Instagram and Facebook changing and stay with the platform, or you leave. And right now, there hasn’t been enough of a different platform that has changed up the space to get people to move over there.

Inadvertently, Eva seems to presage the rapid uptick of TikTok; accordingly, several of our interviewees noted how creators had abandoned Instagram’s oversaturated community for the short-form video platform. Recall, for instance, the earlier mentioned content from Kavianah, who moved from Instagram to TikTok because—in her words—she was not “being seen.” Josie, similarly, noted how many creators saw their visibility decline at the hands of Instagram and subsequently rechanneled their creative energies elsewhere. As she explained during our 2020 interview,

Even to this day, a lot of people are still annoyed at Instagram and have moved over to YouTube and TikTok because [Instagram] is not a sustainable career . . . Instagram is one of the most penalized and probably the least favorite because of just the way [its algorithm] was changed in 2017.

An even greater threat was that the platform on which creators had built their brands would disappear without warning. Naina, for instance, expressed a concern that Instagram could “fail” at any time, leaving users without access to audiences and advertisers. She added, “If you don’t have a backup plan, you’re done. You’re gonna have to work at McDonald’s, because [you don’t have other] experience probably and all you’re gonna be able to say is, ‘Oh, I had 500,000 [Instagram] followers.’”

TikTok creators seemed acutely attuned to this type of precarity, especially given that we conducted interviews at a time when the platform’s future in the United States (where most of our interviewees were located) was unclear. Cooper recalled countless comments from his audience in anticipation of such uncertainty, such as “Oh, I really hope TikTok doesn’t get banned. I’ll miss you.” However, he seemed optimistic about the potential to find another way to reach his audience, saying, “I’m not going anywhere. I’m a human being. I will still exist. I will still make my content. It just won’t be on TikTok.” But while Cooper was certain about his future in content creation, the majority of our interviewees noted how unpredictable variations in the broader industry ecology were frustrating at best, and beleaguering at worst. Like Cooper, they expressed the need to have an alternate plan of action should their platform of choice fail. As Nickolas put it, “I believe TikTok is not very future-proof at its current state. And many creators, if they can’t really make a living, then they might just move on to other platforms.”

As such comments attest, creators believed that developing a cross-platform brand was a way to better withstand the vicissitudes of the platform ecology (see also, Cunningham & Craig, 2019; Glatt, forthcoming). Nolan explained why he was also creating content on sites like YouTube and Instagram:

You’ve probably seen on the news too about the potential TikTok ban, everything happening with the app . . . [It has] given me a wake-up call to try and not focus so heavily on one platform, even though I’m doing so well, and to try and have audience on multiple platforms just in case my account gets banned . . .

While this strategy may be a bulwark against the instability of the platform industry, it places an additional burden on creators by demanding further investments of time, energy, and creative attention.

Platform Precarity: Features and Algorithms

The breakneck pace of evolution is an established feature of the technology sector. Our interviewees, accordingly, reflected on the disruptive nature of platforms’ continuous and oft-unannounced “updates.” While these changes collectively added to workers’ experiences of precarity, changes to platform features were seen as more transparent, whereas, other changes—particularly tweaks to underlying platform algorithms—were more opaque and thus deemed harder to reckon with.

Transparent Changes in Platform Features

Changes in platform features command creators to continually adapt, swapping yesterday’s modus operandi for today’s newest feature rollout. Several Instagrammers reflected on how they had felt compelled to use the “Stories” feature introduced in 2016, which has since changed in functionality. Recalling her initial frustration with the introduction of this feature, Olivia said,

When it [Instagram stories] first came out, I was like, “This is stupid, I’m not doing this, it’s just something else [to do] . . . I don’t need another social media platform or element to use. I already have enough.”

Yet, she added, “[but it] has turned into a habit that I literally do every single morning.” Some interviewees expressed frustration that when a platform did announce a change to their features, communication was often unclear, leaving more questions than answers. Blogger and content creator Neve explained that in 2018, Pinterest advised creators on the best search engine optimization (SEO) practices with the instruction that “the first five pins that you save during the day will have the biggest priority.” She continued,

One week [later], they changed their mind. They removed that line from the “best practices” page. And now about six months later, the best practices page has been completely rewritten, changed, and it doesn’t even exist. Now it’s called “content tips.”

That any platform decision may be abruptly reversed was an additional source of stress and uncertainty for creators. Such capriciousness, Neve concluded, leaves creators “really confused about what Pinterest wants.”

TikTok creators, meanwhile, expressed confusion about changes in the TikTok Creators Fund, a program which—similar to the Creator Marketplace—promises to help members of the community monetize their content. The guidelines are notoriously vague, and interviewees thus offered inconsistent understandings of metric benchmarks. While some noted a qualification of having “10,000 followers or more,” others believed admittance was based on garnering “10,000 views or more in the past 30 days.” As Kavianah explained,

A lot has changed since I first joined in April and a lot of it has to do with that TikTok Creators Fund . . . It’s got people really looking at the content that they [put] up, looking at their analytics, looking at their follower gain because, I mean, you’re getting paid for your [hits] on that, so it’s making everyone really want to make sure that they’re being seen . . .

Together, interviewees seemed to suggest that platform changes sparked widespread uncertainty among members of the creator community, even if these features were more transparent and easy to discern than changes to the algorithmic logic.

Opaque Changes in the Algorithmic Systems

In addition to enacting publicly discernible updates to their features, platform companies update their algorithms in earnest, mostly in ways that are not transparent to users and creators. Our interviewees reflected at length on the role of such notoriously “inscrutable” systems in amplifying the uncertainty of their work. While Angela noted how Instagram’s algorithm “constantly, constantly changes,” Eva explained, “Instagram doesn’t really share what goes into making an algorithm. They tweak it whenever they want, and then sometimes they announce it, sometimes they don’t . . . People just start noticing that things are happening.” To her, such changes were economically motivated: “If they [Instagram] can change the algorithm to help make more money, then they will make that choice.”

Given the opaque nature of such changes, many creators cast them as an incessant source of frustration, particularly as they required additional time and energy to “keep up.” As Grace noted of Instagram’s “constantly changing algorithm”: “it’s a pain in the ass to keep up with.” Then, drawing attention to the labor involved in trying to “game” the algorithm by participating in reciprocal posting groups, Grace told us: “I spend just like a complete hour right after I post . . . liking and commenting on other people’s pictures and responding to comments.” Cooper offered a similar reflection on TikTok’s algorithm:

Something that’s worth noting is the unpredictability of the algorithm on TikTok. It’s not always super consistent, and for some reason, you’ll go through a week or two where your views are half of what your views are in other weeks. [It happens] even if the production of your content—the quality—isn’t changing . . .

Elizabeth summarized, “It seems like nobody really has a firm understanding of how [the algorithm] works or what it means. I think the lack of transparency isn’t really good for anyone in that case.” Nancy, who had experience creating content for YouTube in the past, even opted to abandon that platform entirely because, in her words, “the algorithms there are seemingly impenetrable.”

Some interviewees came to understand the impact of algorithms in the context of perceived punishments, such as when their profile or content was concealed from audiences—making the creator less visible and, correspondingly, less valuable (Bishop, 2019; Caplan & Gillespie, 2020; Petre et al., 2019). Despite the speculation surrounding platform retribution (see, for example, Myers West, 2018), algorithms are widely understood as the initiators of these metric punishments given their perceived role as the technical arbiters of platforms’ community guidelines. As Neve noted, “Pinterest can block a very legit blogger for nothing . . . really for nothing.” Yara, meanwhile, described it as “kind of ridiculous that you have to take these super careful, extreme measures to not get punished by Instagram.” Being “shadowbanned” by a platform—which users understood to mean that the platform was hiding their content without notification—was a source of angst for many of our participants. As Alex explained of TikTok, “[You might notice that] your followers might go down a bit, and your views aren’t as high as they used to be, but you’re never specifically told ‘you’ve been shadowbanned.’” Josie, who maintained a cross-brand presence while also providing consulting services to aspiring creators, shared, “I’ve been shadowbanned probably five times in the past year on four different accounts I launched. The algorithm makes mistakes . . . they sometimes will literally ban you for no reason.” One site where concerns about shadowbanning circulated was in the context of TikTok’s earlier mentioned Creator Fund. After getting accepted into the Fund, Kavianah noted, “It’s not as easy for you to show up on people’s following pages as it was before, [so] you have to work harder to get on the ‘for you’ page.” Alex, too, suggested that participation in the Fund might make creators less visible: “[There’s] definitely a correlation between joining the Creator Fund, and views, and videos, and everything not doing as well.” Yet because these shadowbanning decisions are not transparent, some questioned whether it was even a “real” platform practice. As Colt described,

TikTok is such a random platform. One video will get 5 million views, and the other video will get 200. So, it’s easy to say that you’re shadowbanned, or TikTok doesn’t like you, if all of your videos don’t do that well. But I don’t think [shadowbanning is] that prevalent, if it even does exist.

Although uncertainty shrouded this practice, most interviewees agreed that with something as nebulous as shadowbanning or other algorithmic punishments, workers have few mechanisms for recourse.

Despite this lack of formal recourse, users still tried to ameliorate uncertainties and mitigate the impact of algorithms on their visibility. Elyse indicated that algorithm changes are to be expected and creators must learn to adapt: “I don’t think the algorithm changes can be predicted, nor should they be resisted—in terms of SM [social media] marketing, it’s just a case of being super nimble and quick to change.” Thus, much like ordinary users who rely upon algorithmic imaginaries (Bucher, 2017) or “folk theories” (Eslami et al., 2016; DeVito et al., 2017) to make sense of purportedly “black-boxed” algorithmic systems, cultural producers developed and circulated platform-centric algorithmic theories in earnest. To Neve, Pinterest’s most recent algorithm change reflected the platform’s wish for “content creators to push more fresh content.” She explained, “Instead of re-pinning their popular pins, they want us to create almost every day some new images, which is frustrating in a way . . .” Nickolas, who was in the tech space, told us about a theory that he has been employing on TikTok, “I hear a lot of people say that using music in your videos helps with the algorithm, so TikTok recommends it to more people.”

These folk theories reveal how many creators went beyond merely acknowledging the existence of algorithms, but rather, to fundamentally orienting themselves to anticipate the incessant tweaks and oft-unforeseen updates of algorithmic systems. Angela shared her understanding that “the basis of the Instagram algorithm is really about trying to show people the content that they want to see and keep people on the Instagram platform as much as possible.” Consequently, she continued, “You can kind of think about the different ways that Instagram might do that, and different rules they’re gonna put in place to make that happen.” This perceived obligation to reckon with the volatility of platforms’ algorithmic systems introduces a condition that we have elsewhere described as “algorithmic precarity,” wherein anticipating the impact of these systems becomes a necessary part of the job (Duffy, 2020).

Discussion and Conclusion: Instability, Intensification, and Inequality

We’re kind of all at the whim of whatever these companies want us to do. And sometimes they’re privately owned, sometimes they’re publicly owned companies, but they’re still companies whose main goal is to make money. (Eva)

It has been nearly four decades since media sociologist Todd Gitlin (1983) described TV producers’ fraught attitudes toward industrial ratings systems: the quantifiable figures supplied by Nielsen and the American Research Bureau that promised to “predict the unpredictable” were simultaneously valorized and trivialized (p. 31). This idiom—predicting the unpredictable—remains apt today amid data-driven platforms, predictive algorithms, and the wider forces of platformization that continue to reconfigure the cultural industries (Nieborg & Poell, 2018; Poell et al., in press). Indeed, our interviews with platformized creative laborers—those who use platforms and also, crucially, direct value-generating attention to them—reveal how their experiences are profoundly wrought by such unpredictability. Visibility, as we have shown, is a key vector of such instability; creators vie to boost their likes, views, favorites, and shares, while simultaneously trying to allay the threat of being rendered invisible (see also, Bishop, 2019; Bucher, 2012).

In an effort to more clearly identify the role of in/visibility in the platform-dependent creative economy, we offer an ecological model of the nested precarities of visibility. Through the idea of nestedness, it becomes clear that any form of change—be it the “rupture” of a new algorithm (DeVito et al., 2017; Petre et al., 2019) or a user movement that seems to reshuffle the industry ecology—cannot be understood discretely; rather, precarity is located within wider regimes of instability—much like Russian Matryoshka dolls nested within one another. Working from the outermost layer, we contend that market forces structure the broader ecology of the platform industry, which in turn shapes the decisions of individual platform companies; together, these interconnected forces guide the experiences of individual creator-subjectivities. As the accounts of the Black Lives Matter movement, COVID-19, and politically motivated platform bans attest, unanticipated shifts in wider culture can trigger sweeping changes in the platform ecology as well as more acute changes in a platform’s algorithm. To be sure, we do not claim these examples are paradigmatic (nor are they finite); rather, they are a testament to the fact that cultural products have long been shaped by the wider cultural environs. At the same time, in the age of instant digital connection, such cultural shifts can have immediate impacts that quickly ripple through the platformized creative economy.

In assessing the multiple, interconnected levels on which creators experience precarity, it seems necessary to reiterate that structures of instability have deep roots in the media and cultural industries, where “job insecurity” is an endemic feature (e.g., Gill, 2010; Hesmondhalgh & Baker, 2013; McRobbie, 2016; Neff et al., 2005). Interviewees’ accounts of fickle audience tastes and curious market vicissitudes are not dissimilar to those chronicled in earlier studies of media workers (e.g., Ang, 1991/2006; Gitlin, 1983; Turow, 1997). But the individualized, data-driven nature of cultural production and consumption has exacerbated such metric unpredictability—at least across mainstream creator platforms. The rate of change, too, seems to have accelerated, guided by Silicon Valley’s siren call to “move fast and break things.” As one of our interviewees (Josie) aptly noted of the breakneck pace of algorithmic changes, “By the time TikTok discloses, or anyone figures [the algorithm] out, it changes again. So, it’s really hard to keep up with everything.”

Accordingly, while cultural producers have long sought to appease audiences and advertisers, today’s platformized creative laborers are also—and, perhaps, primarily—beholden to platform companies—those that hold a perverse level of power in contemporary culture and society (Van Dijck et al., 2018). Interviewees’ efforts to curry favor with platforms by, for instance, hitting metric benchmarks or engaging in actions that will (hopefully) stave off punishment are not merely expressions of professional self-interest; instead, they reveal how creators are enlistees in the broader systems of platform capitalism (Srnicek, 2017). Eva’s above-mentioned comment about being “at the whim of whatever these companies want us to do” suggests that, when it comes to their understanding of platforms, creators are less blinkered; more duty-bound.

Here, it seems important to call attention to the algorithmic systems that govern visibility on so-called “mega-platforms” like Google (YouTube) and Facebook (Instagram). Their algorithmic systems—as considerable research has helped to lay bare (e.g., Bishop, 2020; Gillespie, 2018; Noble, 2018)—are far from “neutral” or “fair.” For instance, in their analysis of YouTubers’ experiences of demonetization, Caplan and Gillespie (2020) describe how the system of governance is shaped by “different users, different sets of rules, different material resources and opportunities, and different procedural protections when content is demonetized . . . [all of which are] different from the expectations of fairness creators bring to the table” (p. 2). Our findings, too, testify to the profound unevenness in the creator economy’s system of “winners” and “losers.” To be sure, some creators benefit from this lopsided system of accruing attention, especially when an unexpected boost in visibility generates career windfall. But most creators experience visibility as a source of not just promise, but also precarity. Therein lies the rub: for as much as the ideal of visibility structures the sprawling creator economy, the systems that enable or constrain it are—paradoxically—invisible. The newly formed cottage industry selling “tricks” and “hacks” to game the algorithm (Bishop, 2020; Chan, 2019) highlights the mystery that shrouds the pursuit of visibility. Accordingly, it is important to consider who has the time, energy, and financial resources to devote to learning strategies of growth and visibility in hopes of leveraging what Klawitter and Hargittai (2018) describe as “algorithmic literacies” (see also, Cotter, 2019). Put differently, platforms’ systems are inscrutable (Rosenblat & Stark, 2016), yes; but perhaps it is more apt to frame them as more or less scrutable.

In closing, we call for additional insight into the allocations and rewards of visibility in platformized creative labor. While we have focused on those social media sites commonly understood as “mainstream” platforms, recent work on DIY musical artists reveals marked disparities both within and across categories of “consumer-oriented platforms” (e.g., Spotify) and “producer-oriented” platforms (e.g., SoundCloud and Bandcamp) (Hesmondhalgh et al., 2019; Jones, 2020). Meanwhile, the kind of algorithmic punishments we have noted warrant further attention given accounts that they disproportionately impact women, racial/ethnic minorities, people with disabilities, and the LGBTQ + community (Bishop, 2020; Glatt, forthcoming). While only a handful of our interview participants discussed social inequalities—largely in the context of the (in)visibility of Black creators on TikTok—reports in the media show how these rewards and punishments map onto existing structures of marginalization (Frier, 2021; Parham, 2019). For members of these groups, the precarities described above are likely further exacerbated, and we see a focus on their experiences as a key avenue for future research in this space. After all, any nested system of precarities is ultimately embedded within a social structure that is far from stable—or equitable.