Desperation and Datalogix: Facebook Six Months after its IPO

Facebook began trading publicly on NASDAQ nearly six months ago, on May 18, 2012, at the initial offering price of $38 a share. Despite unprecedented anticipation and early trading volume, the company’s share price soon declined, and it currently hovers around the $19 mark. One of the reasons for this decline stems from an issue that plagues all advertiser-supported media: how does the company know if ads actually result in consumer spending? In addressing this concern of Facebook shareholders, the site has recently partnered with data management firm Datalogix in order to bolster the link between Facebook ads and in-store purchasing.

From a political economy perspective, Facebook’s partnership with Datalogix also bolsters the site’s commodification of users, where Dallas Smythe’s “audience commodity” concept (1977; 1981) used to describe television viewers might be updated in the new context of Facebook/Datalogix data collection and social media marketing. Below are few of the ways in which the audience commodity framework could be applied to this partnership.

The audience as users/consumers: Even more so than television audience members, Facebook users function as workers who look at (and click on, although clicks are apparently less important than originally thought) advertisements but are also, crucially, suppliers of personal information and producers of content – particularly on a platform like Facebook, where members must disclose personal information at the outset in order to use the site, and then provide an overwhelming amount of site content as a function of their participation. Yet from the perspective of Datalogix, users are consumers in a more traditional sense – offline user buying habits are now being connected to online profiles as a way of gauging how effective Facebook ads are at actually generating in-store purchases. Aside from the potential privacy implications of this practice, it serves to amplify the commodification of Facebook users who are more and more explicitly addressed as market share and described in terms of “reach.”

Programming as site content: Content is still important in this scenario – Smythe famously characterized television programming as the “free lunch” that would draw audiences to watch advertisements. But content must also provide the right environment for advertisements (Jhally & Livant 1986: 140). Outside of certain instances of unacceptable content, Facebook users enjoy relative freedom to add whatever they wish to their profiles and communicative exchanges. Yet the site does employ design elements that encourage an environment conducive to advertising – especially the “like” button, which allows users to “like” any content posted to site.

The like button appears underneath wall posts, photos, and videos, and also underneath sponsored posts and sidebar advertisements. Moreover, the like button extends beyond Facebook itself to be integrated across many different websites for news content, e-commerce, and multimedia. The prevalence of the like button, along with its ease of use – just pressing “like” as opposed to formulating and writing a response in the comment box – fosters an environment of positive endorsement. There is no option to “dislike” or even to “not care”; liking something effectively feeds into the commodification structure of the site as part of users’ labour of endorsement that takes place in a positive atmosphere. So for example, liking a Facebook friend’s baby photos can lead to targeted ads for baby-related products, especially for female Facebook users who are also listed as married.

Ratings systems as data-mining: The data to be gleaned from Facebook users represents both a quantitative and qualitative change from the data gathered through television ratings systems. Nielsen’s “people meter” accounted for what TV programs each family member was watching and for how long; Facebook compiles personal information about a user’s identity, location, work, and education, along with her or his preferences in books, music, film, and activities, as articulated among an entire network of “friends,” situated within the user’s photographs, wall posts, messages, and event pages, and complemented with behavioural data on her or his “likes,” clicks on hyperlinks or advertisements, and search terms. And because all of these pieces of information are transmitted digitally, they leave traces that are generated automatically and persist potentially indefinitely. Such an exponential expansion in the quantity, quality, and durability of user data means that commodification on social media sites takes place at an accelerated rate and happens immanently to participation in these sites. When compared to this context of “big data” and increasingly sophisticated technologies for harnessing that data through the valorization of online surveillance, the television ratings systems of the past appear clumsy and quaint. But at the same time, the ways in which television audiences were segmented was not strictly empirical, as Eileen Meehan (1984; 1986; 2002) points out in her gender critique of the audience commodity, but was a function of ratings companies’ creation of particular audience groups. Obvious gender stereotypes persist in Datalogix’s “syndicated segments,” especially in groups like “Soccer Moms,” “Green Consumers,” and “Sports Fans”. In turn, the categorizations of such groups based on gender (and also other stereotypical features of class, race, ethnicity, and age) function as a kind of discrimination by assigning differential value to these different target markets.

Narrowcasting as target marketing: Organizing user groups into target markets finds its precedent in the narrowcasting strategies of cable television in the 1980s, where the gap between programming and advertising content was increasingly blurred by the targeting of both toward specific audience groups. Yet different audience groups were not created equally in this new television landscape, where the more valuable audiences – usually comprised of white, higher-income, male viewers – were the ones who were measured by ratings meters and thus the ones who set the benchmark market value for audience power (Meehan 1990: 132). Target marketing does not just reflect user desires, it produces them in ways that are differential according to already existing structures of privilege. For example, the user profiles that underlie the targeting of specific content to specific user groups also result in price and marketing discrimination. In a process known as “weblining,” data-mining firms like Datalogix organize the information collected through Facebook into profile groups based on income, such as “Homeowner Status,” “Presence of Children,” and “Credit Card Buyers”. In turn, these groups are presented with differential pricing or different products, based on their perceived spending power. This unfair treatment tends to involve discrimination against lower income groups, non-whites, and women, and this discrimination is particularly troubling when it extends beyond the ability to make consumer choices into the ability to access information in the public sphere.

While there are certain qualifications to made around the audience commodity when transposed from the context of television to that of social media – where audience members have become users, programming has become site content, ratings systems have become data-mining, and narrowcasting has become target marketing – the continuities between these media environments are significant. The widespread popularity of television forms a necessary precondition for the advertising environment of the web, where looking at advertisements is a long-quotidian practice and advertising is naturalized as the most efficient profit model for delivering “free” programming. This continuity between television and social media highlights how, while social media platforms often present themselves as general “means of survival,” their use is designed to serve their own profit interests. Nowhere is this designation more clear than in Facebook’s recent partnership with Datalogix, as a desperate move during a telling time in the company’s publicly traded existence.

Note: This post is excerpted and adapted from my forthcoming chapter on social media and the gendered commodity audience in The Routledge Handbook of Media and Gender (eds. Lisa McLaughlin, Linda Steiner, and Cynthia Carter, 2013).

References

Jhally, S. and Livant, B. (1986) ‘Watching as working: the valorization of audience consciousness’, Journal of Communication 36(3): 124–143.

Meehan, E. (1984) ‘Ratings and the institutional approach: a third answer to the commodity question’, Critical Studies in Mass Communication 1(2): 216–225.

Meehan, E. (1986) ‘Conceptualizing culture as commodity: the problem of television’, Critical Studies in Mass Communication 3: 448–457.

Meehan, E. (1990) ‘Why we don’t count: the commodity audience’, in P. Mellencamp (ed.) Logics of Television: Essays in Cultural Criticism, Bloomington: Indiana University Press.

Meehan, E. (2002) ‘Gendering the commodity audience: critical media research, feminism, and political economy’, in E. Meehan and E. Riordan (eds.) Sex & Money: Feminism and Political Economy in the Media, Minneapolis: University of Minnesota Press.

Smythe, D.W. (1977) ‘Communications: blindspot of western Marxism’, Canadian Journal of Political and Social Theory 1: 1–27.

Smythe, D.W. (1981) Dependency Road: Communications, Capitalism, Consciousness and Canada, Norwood NJ: Ablex.