The human capacity to communicate is at the core of our sociality. As a species, we are social to the extent that we communicate (Levinson and Enfield 2006). This truism has acquired a special meaning in the age of the internet, when a great deal of our communication is mediated by computing platforms such as email, social media, electronic commerce sites, and all the rest. The historical novelty is in the integration of multiple forms of communication into the composite and interactive medium of the internet. The bigger novelty, however, is in the fact that a large portion of this medium is now owned and operated by private corporations that have emerged as global powers in the last couple decades. A few numbers can put this into perspective.
From December 2008 to December 2013, the number of users on Facebook went from 140 million to more than one billion. During this same period, Facebook’s revenue rose by about 1,300 percent, with ad revenue rising from $300 million to above $4 billion. Facebook’s filing with the Securities and Exchange Commission in 2012 indicated that “the increase in ads delivered was driven primarily by user growth” (Facebook 2012, p. 50). LinkedIn, similarly, has enjoyed exponential growth (300+ million members as of February 2015), but most of its revenue comes from “a small group of recruiters looking for talent,” along with advertising income and fees for its Premium Membership1. Twitter, with an income of $1.3 billion in 2014, has banked on our desire for validation: “What you say on Twitter might be seen around the world instantly.” Schiller notes that in 2013, Google “announced to users of its YouTube, Google Play, and other services that their photos, profiles, comments, and rankings might be deployed in constructing advertising endorsements across the two million Web sites served by its display-advertising network, sites that are viewed by an estimated one billion people” (2014, p. 173). Validation comes at the price of corporate ownership of data we choose to share, such as photos and personal news, but also extends to the manufacture of data we have not shared, i.e., data constructed about us, such as rankings and classifications that use the raw material of our posts and clicks to tell stories about us that are not of our own making.
Many startups have also emerged in the last few years with business models that monetize online analytics—i.e., the data extracted from patterns of communication on social media and other platforms. Pulling together data from Facebook, LinkedIn, Twitter, and other sites, a service such as Klout, for instance, provides a score between 1 and 100 for an individual’s “influence” in terms of the size of their network, its content, and the number of users interacting with the content. Another company, Acxiom, describes one of its lines of business as global media companies “who grow revenue and market with easy-to-sell, high-value audiences, using the best possible consumer data and accurate, permissible data matching services in a privacy compliant manner” [emphasis added].2 Providing information to corporations, nonprofits, political organizations, and government agencies, Acxiom reassures us—citizens, consumers, donors—that we “get more relevant advertising and special promotions and [that this] reduces the chance that you’ll become an identity fraud victim.” A few lines below this assurance it is mentioned that this information might, however, include “sensitive information such as the Social Security number.”3
The direct relationship between the growing number of users and the expansion of these large and small companies is a well-accepted verity among observers—indeed, some seem to plainly state that heteromated labor extracts significant value (without, of course, using the term). According to the VP for Research of the technology consulting firm Gartner, Inc., for instance, “Facebook’s nearly one billion users have become the largest unpaid workforce in history” (Laney 2012, emphasis added). Writers who analyze this workforce from a Marxist perspective consider the mechanism of value extraction to be the same as the traditional capitalist exploitation of surplus value. Fuchs (2011), for example, argues that users of social media are infinitely exploited because they contribute their labor-time without any reimbursement. This argument is based on the earlier theory of “audience commodity,” according to which users of media (e.g., TV watchers) are sold as a commodity to advertising companies and their clients (Smythe 1981). Other scholars find the mechanism in the capability of companies to aggregate the “affective investments” of users (Arvidsson and Colleoni 2012), and still others the commodification of personal data (Andrejevic 2015; see Profitt, Ekbia, and McDowell 2015 for a compendium of these views).
5.2 The Networked World: Mobility, Separation, and the Desire to Connect
We consider the mechanism of such value extraction as a form of heteromation that allows both securing value and making it invisible through data production. The interposal of privately owned computer networks into human communication has turned interpersonal communication into a kind of labor, the economic benefits of which go to corporations. In the “networked world,” people are valued to the extent that they can forge new links, and, to some extent, remain distrustful of (and, hence, disloyal to) pre-established structures and institutions. This is a world where forging relationships can be a source of benefit (“social capital”), and where people’s value derives in part from their degree of mobility. Mobility, however, comes at a cost—most obviously, of losing one’s footing in community, with the attendant predicaments of separation and status anxiety we discussed in chapter 4. These predicaments, reinforced by the uncertainties of unstable working environments, provide a powerful incentive for staying in touch and the underlying push toward social media.
Web 2.0 technologies respond to this situation by providing a space of interactivity that purports to fill the gap of separation. The push of predicaments is complemented by the pull of technology, which pragmatically and programmatically enables remote and persistent connection. In this fashion, interaction in the networked world brings a diverse set of actors—manufacturers of networking technologies, software development houses, ISPs, and social networking sites, along with “users”—into contact with each other as nodes in the same network. In so doing, it turns communication into a computable activity that produces rewards for the individual and economic value for the other actors.
5.3 Social Data: Value Extraction through Communicative Labor
Web 2.0 computing platforms are, by and large, dedicated to the production of social data—that is, content generated by users through online interaction and communication. Interacting with others in these environments through “tagging,” “sharing,” “liking,” “upvoting,” and so on, involves actions that produce, as a by-product, different types of social data: descriptive (e.g., individual profile data), behavioral (e.g., “liking”), and user-generated (e.g., photos, videos, comments). Social data is different from other kinds of data in its character and in how it is created, and both of these differences are important for understanding its economic value.
5.3.1 What Is Social Data?
In terms of its content, social data is an innovative concept. Unlike sociodemographic data that have been traditionally collected through the institutionally sanctioned expertise of statisticians, demographers, accountants, marketers, and epidemiologists, social data is collected from activity in online environments. Rather than being about social behavior, in other words, it is produced through social behavior, acquiring a fine granularity that was absent from the aggregate statistical data of yesteryear. Facebook’s EdgeRank algorithm, for instance, analyzes around 100,000 parameters in order to push the right type of content (what to watch, whom to add, and so on) to a user. Social data is, therefore, also personal in at least two senses—first, in the way it is organized, presented, and “personalized” (van Dijk 2013), and then in the sense that its effects are fed back to the individual through algorithmic analysis and data mining (Andrejevic 2015).
The transition of personal profiles on Facebook from collections of data items to the narrative structure of a Timeline highlights the first point. Through a retroactive chronological construction of individual life stories and milestones (births, graduations, weddings, and so on), the Timeline provides an organized, digital version of the venerable “shoebox,” from which selected bits and pieces of one’s life are compiled and displayed for the world to see. At the same time, the transition has enabled new techniques of pushing data to users—e.g., through PageInsight data that allows marketers to obtain real-time analytics about user behaviors and preferences. As van Dijk (2013) argues, online platforms’ advocacy of a unique online identity for users derives from their vested economic interest in this aspect of personalization.4
This business concern also relates to the other personal aspect of social data—namely, its capacity to be fed back to the individual in the form of ads and recommendations, as well as the distribution of opportunities such as employment.5 This capacity is behind the idea of “data as asset,” famously announced at the World Economic Forum in 2012 as a maxim of contemporary capitalism. This idea underlies the business model of companies such as Acxiom, which, with major clients such as United Airlines, Macy’s, and Epson, boasts data products such as “Audience Propensities”:
While other big data analytics focus on what was, Audience Propensities focuses on what is likely to be. The result is more effective marketing, higher profitability and a superior customer experience.6
The statement speaks truth to a historical fact about the need of capitalism to speed up the circulation time of products to increase profits. Marx noted a long time ago: “The main means of cutting circulation time has been improved communications” (Marx 1990, vol. 3, 164). The mediating role of commercial media is, therefore, not new from the perspective of market efficiency (Comor 2015), but what is new is the capability to forecast “what is likely to be,” as limited and flawed as this might turn out to be (Ekbia, Nardi, and Šabanović 2015). As Burdon and Andrejevic (in press) note, users are often unaware of the full scope of the uses of their data. They are unaware that they have consented to its collection and usage. For instance, Burdon and Andrejevic studied technically savvy users, finding disconnects between corporate data collection practices and people’s perceptions of those practices. The authors noticed, for instance, a tendency on the part of their study participants to consider the high volume of data collected as providing anonymity—a tendency reinforced by statements such as Google’s: “no human reads your email” (Webb 2004). Burdon and Andrejevic observed that:
While it is true that big data firms such as Google may not be interested in particular individuals per se, it is certainly interested in collecting their data in order to aggregate it with information from other users as a means of more effectively managing and manipulating them—at the individual level. In other words, both of the following claims can be true simultaneously: (a) that Google is not interested in particular individuals but that (b) it will nonetheless collect and store detailed information about users in order to more effectively tailor advertising and other forms of content to them, and indeed to the general population of users. In that sense, lack of interest does not mean lack of impact.
Social data, while maintaining an innocuously impersonal appearance, acquires a very personal character at the same time.
5.3.2 How Is Social Data Collected?
Social data is also different in how it is obtained. What might seem to be “natural” forms of communication are, in fact, carefully tailored modes of interaction intended to generate data footprints. By programming user participation so as to make interactions computable, Web 2.0 platforms collect data in ways that are distinct from the generation of data through automated technologies of monitoring, data tracking, and recording. This presents an interesting instance of how heteromation trumps automation, surpassing it when possible and subsuming it when needed.
The standardization of the interface is the first step in this direction. By providing specific placeholders (e.g., for photos of particular life events), for instance, the Facebook Timeline cues the user as to the type of content they should include on their profile. In this manner, these platforms create and promote a special kind of user—a user who “shares” (John 2012; Shilton 2016).
The standard interaction templates go further, formalizing everything on the interface (users, photos, posts, comments) as “objects” related to each other through pre-established actions such as “following,” “sharing,” and “clicking.” As Alaimo and Kallinikos argue, “behavior-related data is thereby obtained by encoding social interaction qua action into a data-connection between objects.” Rather than simply “capturing” data, in other words, “social media encodes, formalizes, and constantly manipulates data” to produce a new kind of sociality. What makes these encodings novel is that data, rather than top-down categories based on established taxonomies of education, income, demography, and so on (e.g., a male applicant with a college degree and annual income of $60,000), dynamically put people in transient and invisible buckets such as “a good borrower,” “a bad insurance payer,” “a promising job applicant,” “an avid reader of mystery novels.” The translation of these dynamic categories into specific business practices of economic value, such as employee recruitment, product placement, product development, and market strategy development, provides value to corporations that buy the data from social media companies and other operators of Web 2.0 platforms. At the same time, however, the complexity of the analytic processes that generate the categories makes them opaque to the user.
5.4 The Invisible Value of Social Data
The participation of the masses in networks where communicative labor takes place presents an ideal person as someone who is constantly on the move, not only geographically (between places, projects, and political boundaries), but also socially (between people, communities, and organizations) and cognitively (between ideas, habits, and cultures) (Boltanski and Chiapello 2005). The majority of people who participate in these environments find tremendous value in the capability of such technologies for staying informed and in touch, forging new connections, and maintaining old ones. Most often, they do not think of their activities in terms of economic value, to themselves or to others. Such inattention is partly because of the convenience and feelings of connection that networking technologies provide, and partly because of the invisibility of the mechanisms. The contribution, as Boltanski and Chiapello have noted about computer-mediated networks in general, “must at once possess limited visibility … and have meager value … while contributing to [an enterprise’s] enrichment” (ibid., p. 361). Modern persons find value in the use of social media technology through the rewards that it provides, thereby generating “enrichment” for the companies that provide the services.
5.4.1 Self-Validation as Reward
The precarity of current socioeconomic circumstances has increased status anxiety and the desire for affirmation from peers and others. This desire is met by the provision of the many microvalidation mechanisms provided by social networking technologies: “like” on Facebook, “follow” on Twitter, “endorse” on LinkedIn, “upvote” on Reddit, and so forth. Despite differences in style—for instance, between the “friendly” character of Facebook likes and the professional look of LinkedIn endorsements—these mechanisms share the purpose of providing peer and social validation. Displacing the common mechanism of “examination,” which Foucault identified as a key disciplinary technique of modern societies, microvalidations draw on a combination of memory and emotion to encourage the conscious release and sharing of personal information (van Dijk 2013).
Professionally oriented sites such as LinkedIn and ResearchGate use the same mechanism but with an extra layer of formality, providing users with “profile stats,” an “RG score,” and other similar measures. The conscious and voluntary appearance of these behaviors has led some commentators to conclude that:
[N]orms are changing, with confidentiality giving way to openness. Participating in YouTube … Flickr, and other elements of modern digital society means giving up some privacy, yet millions of people are willing to make that trade-off every day. (McCullagh 2010)
The line between open and confidential or conscious and unconscious contribution, however, becomes murky if we take into account the effect of the delicate mechanisms of incitement employed by the sites. The People You May Know (PYMK) feature on LinkedIn, for instance, nudges people toward new connections, which are then reinforced by “invitations” from people one might not even know. While a person’s response to such invitations seems conscious, refusal may feel awkward, given that turning down an invitation always carries a cost in guilt or feelings of lost opportunity.
5.4.2 Self-Promotion as Reward
The microvalidations offered by Web 2.0 technologies are enhanced by the provision of a space for self-promotion (a boon to all of us in a world of precarity in employment). The working of this mechanism is most visible for people with celebrity status (pop artists, actors, politicians) who depend on staying in front of people for their livelihoods, and have found in this space new ways to enhance and promote their image and to reach out constantly to their audiences. A wide range of such individuals, from Madonna to the Pope to the leader of ISIS to Malala Yousefzai, makes extensive use of these media to deliver their messages, mobilize supporters, and raise issues. But the average user can also employ the same mechanism to establish an intended image as a potential employee, connection, lover, friend, family-member-who-stays-in-touch. The “social capital” that is garnered can sometimes be further leveraged to acquire economic benefits. Some sites have, in fact, introduced incentivizing techniques—e.g., individuals with high-traffic profiles can post promotional materials on their Facebook walls on behalf of companies. Services such as Klout formalize this practice through aggregate analytics—i.e., by integrating individual information across social media platforms. The overall effect is the ongoing activation of the “cultural circuit of capitalism” (Thrift 2005, p.6), kept alive by the vast material infrastructure of data analytics, networks, and marketing techniques. In such an environment, personal tastes, thoughts, and profiles become commodities (Beer 2009).
5.5 On the “Fairness” of Communicative Technologies
From a heteromation perspective, the activities of users in social media environments count as labor to the extent that they (1) replace or support the work of paid employees (human resources, market researchers, product developers, corporate strategists, and others); (2) contribute to the bottom line (profit) of corporations; and (3) are not compensated financially in direct or indirect ways, or are compensated meagerly. The average person, separated from communities and individuals, driven by the desire to connect, willingly participates in these platforms to benefit from their networking capabilities. In so doing, they provide a trove of data that can then be used to structure their choices, regulate their behaviors, and provide or deprive them of opportunities, as the case may be. This arrangement might seem “fair,” given the rewards the platforms provide for the average user, but is it? While this is a legitimate question from a moral perspective, it occludes the workings of the system, turning the question of human labor into a taken-for-granted supposition rather than as an object of inquiry. Our argument here seeks to avoid this trap, and the perspective of heteromation provides a way out.
This perspective brings out the asymmetries of the participation in networks as an uneven playing field. Participation enables individuals to connect, while at the same time depriving them of the “network-making power” of corporations, businesses, and the state (Castells 2009). Only a small number are able to leverage the capabilities of the networks for the purpose of data collection, turning data into a valuable asset. This group largely involves the nouveau riche of the internet age—that is, the founders of high-tech and social media companies that trade personal data (Forbes 2013).
Conceptualizing user contribution to online platforms as a form of labor highlights its invisible character, revealing the mechanisms of reward that drive people to consciously or unconsciously share personal information and other content. To blame people for the lapse of privacy or to applaud them for their openness tends to obscure these mechanisms, putting the blame or the credit on the individual. The oft-forgotten fact is that human beings, while enjoying a certain degree of choice and agency, are largely in the grip of social and personal predicaments, the impact of which is beyond their control. Meanwhile, the kind of self that emerges from the trade of data troves is, more often than not, an anxious, insecure, and uncertain self, vulnerable to the whims of data merchants who are on the hunt for “easy-to-sell audiences.”
5.6 The Paradox of Mediation
Ferguson (2013) wrote of the tragic heroism of the lone individual facing the enormity of a continent, or a society, or a betrayal. Our loneliness, by comparison, lacks such grandeur. It is assuaged by a succession of fleeting moments of connection that momentarily quiet anxieties and insecurities. Within the difficulty of finding genuine connection online (or off), we accept “next-best-thing” substitutes. These substitutes are evident in, for example, services such as Google’s Let’s Play (LP) YouTube videos.
Let’s Play is currently the most popular content on YouTube, with the bulk of YouTube’s top channels affiliated with it in some way. Let’s Play, and other amateur YouTube channels, are quickly overcoming mainstream media in viewership numbers, with millions of hours of content watched daily. The most popular YouTuber, PewDiePie, is an LPer with over 40 million subscribers. He earns millions of dollars a year in ad revenue.
Figure 5.1 PewDiePie
Most LP creators, however, are ordinary people with only a few subscribers. (We discuss the bifurcation of networks into a small number of highly visible, successful participants alongside masses of ordinary participants in chapters 9 and 10.) LP viewers spend many hours a day watching recorded videos of (mostly) ordinary people playing video games in “playthroughs.” Creators narrate their game action with a running stream of self-commentary and humorous banter. Creators themselves are also viewers. In an interview study of LP creators, one interviewee explained why he watched:
I used to play games with friends when I was younger, but now I’m in college and my friends have moved away. Watching these people play games from my childhood and laughing and joking about it reminds me of that, and I can relive that through the internet. (Major and Nardi 2016)
Here, separation from childhood friends motivated viewing. LP creators tap into the angst of others’ separation by designing video performances with speech, gestures, and settings crafted to project “authenticity” and “genuineness” (ibid.). The videos explicitly present the creator as the viewer’s friend. One creator explained: “You’ve succeeded if your audience thinks of you as the friend next door” (ibid.). Viewers interviewed in the study stated they liked watching “just guys like me” and the “extremely down to earth, just regular people” found in LP videos. (ibid.)
While Facebook channels our communications by means of sophisticated algorithms written by PhDs, the amateurs over at Let’s Play have taken on the possibly more difficult task of putting on a show in which they appear to like you! Creators report that most people who subscribe to and watch their channels are lonely or seeking connection with someone who shares their gaming interests (ibid.). The creator quoted above who guided his audience to think of him as the friend next door, regularly watched several LP channels himself. He spoke of the wistful, illusory quality of the experience:
Often times I myself imagine myself having a great time playing [the popular online game] Minecraft with CaptianSparklez [sic] or Northern Lion [two prominent Minecraft LPers]. But since you can’t actually do that, watching them play and make jokes is the next best thing.
He reports his experience without irony, cheerfully accepting its virtual, “next best thing” nature, much as the Yelp contributor mentioned in chapter 3 accepted Yelping alone on Friday night. People readily seek experience that seems, on its face, a little forlorn.
But we cannot dismiss this activity as mere spectacle, as passive viewing. Let’s Play viewers are watching videos in which they are intimately familiar with the activity through their own participation in the games. And there is, crucially, some interaction with creators. Viewers may receive creator responses to comments they have posted on an LP channel, or responses to replies they have posted on a creator’s Twitter account. Creators may acknowledge comments in a “shoutout” in a weekly vlog, a quite public airing of the connection between viewer and creator. Occasionally viewers even get to play with creators. Some creators devote episodes to playing games with fans whom they recruit on Twitter or in other social media. For viewers, this interaction is the pinnacle of Let’s Play experience (ibid.), a chance to directly participate with the person they have come to think of as a friend. Several means of communication, though often infrequent and attenuated, bring viewers and creators together.
Other spectacles, by contrast, keep viewer and author/performer/creator strictly apart. Football, for example, a game played by only a small minority of men and by (essentially) no women, is nonetheless watched on television by millions of men and women. Football players are definitely not the friend next door, nor would they portray themselves as such. They certainly do not invite fans to play televised football games with them! The participatory nature of the internet, and its copious channels of communication, refuse the spectacle—viewers are swept into the (heteromated) action through direct and indirect interactions with creators. As noted in chapter 2, heteromation, and contemporary capitalism more broadly, work by inclusion, bringing the masses into the fold through varied forms of participation.
A central paradox of Let’s Play is the odd sense of intimacy creators manage to establish with complete strangers. Creators are not the other-worldly Hollywood gods and goddesses of the era of massification, to be placed on pedestals and admired from afar. On the contrary, LP creators’ appeal lies in appearing “authentic” and “genuine,” even to the point of reminding viewers of childhood friends. If creators cannot project these qualities of authenticity, genuineness, and simple friendliness, they are doomed. One creator explained:
Being genuine is the most important thing you can do. If you aren’t genuine, they [the viewers] will know it, and they’ll leave you for somebody else. (ibid.)
The threat of being left for somebody else intensifies as creators begin to attain higher subscriber numbers. Extending Let’s Play-style friendship to lonely people becomes more difficult when creators cannot answer every comment or respond to every tweet as they did when subscriber size was small. Creators must manufacture “authenticity.” Various manipulations come into play, such as the use of production techniques that make the videos appear amateurish. Ad revenue may be rolling in, but video content must not exhibit too much polish, even if the creator has moved on to using professional staff and high-end equipment. Unlike the perfection of, for example, the football replay, LPers may intentionally use low-quality techniques of audio mixing or footage editing, with scenes cutting off too early or too late, or images intentionally cropped poorly. One interviewee said:
You want your content to look like you care, but you don’t want it to look too good. Over-produced, over-edited stuff with what is obviously a team behind it seems less genuine. Once you’re using boom mics and setting up light screens, how are you any different from IGN or Gamestop [professional gaming companies with YouTube channels]? You want to get as high quality as you can, but still within the scope of being amateur. (ibid.)
Even at the apex of Let’s Play success, creators labor to crank out authenticity. For example, PewDiePie yells into his microphone in exaggeration when playing a game that might be scary or tense. Rather than formatting the audio to remove the static and popping that occurs when going beyond his microphone’s volume threshold, these production errors are often retained (ibid.)
Another technique to keep the audience interested is disclosure of highly personal information. One creator said:
I can understand why people get this weird feeling that they actually know you, when they’re just watching your videos. First, you put out so much content, they’re just way exposed to you. Second … you relate some personal stuff on there. You sometimes say stuff you wouldn’t even say to your parents or real life friends, but here you are exposing yourself online. It’s what people come to watch—that connection. (ibid.)
Both creators and viewers buy into a collective fiction of authenticity and connection. The audience knows that successful YouTubers employ teams of video editors and production managers, yet the suspension of disbelief that the videos are still somehow “amateur” is carried forward.
YouTube Let’s Play videos leverage the heteromated labor of viewers and video creators, generating value for Google in millions of hours watched. A small number of LP creators turn their labor into compensated work, and a few become rich. Creators survive day to day on a treadmill of popularity they know they might easily fall off, well aware that their viewers are only too ready to leave them for somebody else if their performances fail to deliver a sense of connection. The paradox of spending hours a day watching videos when one is lonely, or separated from others with shared interests, instead of seeking face-to-face interaction, all the while understanding the manufactured nature of the mediated “connection” and “authenticity” does not bear easy explanation. Nonetheless, we can observe the outcome of the predicament of separation in the economic value generated through media such as Let’s Play.
