Target Markets and Logistical Management

Lindsay Weinberg


Cultural studies scholarship has been at the forefront of analyzing the transformations in culture and political economy underpinning the transition to post-Fordist information societies. In particular, cultural studies has explored the tension in digital culture between forms of online user empowerment such as social media and online community building, which figure the user as both producer and consumer, with the ways the exploitation of user data and digital labor perpetuate inequality. This article works to bring cultural studies scholarship into conversation with the critical study of logistics and supply chain management, with a focus on illustrating how precisely target advertising and the surveillance of user behavior contribute to the capitalist project of risk management and labor exploitation.

Critical studies on logistics and supply chain management often focus on the transformations in the organization of labor that result from an emphasis on the circulation of commodities. Anna Tsing defines supply chain capitalism as

commodity chains based on subcontracting, outsourcing, and allied arrangements in which the autonomy of component enterprises is legally established even as the enterprises are disciplined within the chain as a whole. Such supply chains link ostensibly independent entrepreneurs, making it possible for commodity processes to span the globe. Labor, nature, and capital are mobilized in fragmented but linked economic niches.1

For Tsing, supply chain management through the logistical coordination of labor, transport, and consumer demand has allowed capitalism to efficiently exploit the “enhanced mobility of labor and the economic and political vulnerabilities created by recent forms of imperialism and histories of global war.”2 Similarly, Deborah Cowen is concerned with what she describes as the “new framework of security—supply chain security” that capital’s emphasis on circulation necessitates: a framework which “relies on a range of new forms of transnational regulation, border management, data collection, surveillance, and labor discipline, as well as naval missions and aerial bombing.”3 Target marketing—the segmentation of consumers according to their demographic data, buying habits, preferences, and/or location for the purposes of advertising—and practices of leisure-time surveillance are not generally framed as part of the shift in capital’s emphasis on circulation. However, if part of logistical management is about the displacement of labor to the underdeveloped world, it is equally about monitoring circulation and demand in the overdeveloped.

This paper argues that situating target marketing as a technology of logistical management emphasizes the importance of information in not only intensifying and maximizing the productivity of supply chains and reducing labor costs, but also increasing the likelihood of a return on capitalist investment through the management of market choices. The paper begins by analyzing how target marketing operates as a technology of risk management via consumer surveillance. I then frame target marketing as part of the historical trajectory of the revolution in control described by James R. Beniger. I demonstrate how target marketing provides valuable point-of-sale and point-of-interaction insights, and platform providers can wield this information not only to control prices and allocate advertisements, but also to manage distribution and arbitrage the labor market. Rather than conceptualizing the production of user data as a form of labor in the context of target marketing, I argue that the labor theory of value is being misapplied to the conditions of leisure-time surveillance and data aggregation essential to target marketing. The explanation of user attentiveness as a site of labor not only disregards the relationship between value and time, but also tends to collapse distinctions between the workday and leisure-time surveillance in ways that mystify the differences in how capitalism exercises control over subjects. I then provide a close reading of a fulfillment center exposé in order to examine precisely how the information produced during leisure-time surveillance can impact the conditions of fulfillment center labor. Target marketing, then, is part of a larger logistical apparatus that aggregates data for the purposes of assigning risk, differentiating prices, and managing supply chains and labor costs. It equally reinforces biases and discriminatory practices prevalent in financial institutions in order to maximize profit through the aggregation of data produced by users during seemingly innocuous acts of consumption and online attentiveness.

Risk and Differential Pricing

Target marketing is not simply a means of presenting users with the “best” options and choices on the market or the most relevant information to their preferences and desires. Target marketing works by tracking consumer behavior, preferences, likes, cursor hovers, purchasing habits, location, and any other useful information directly provided during user registration, captured by cookies, and/or purchased from third-party data collection services in order to determine the presentation of advertisements and, depending on the company, manage supply chains. For instance, Amazon uses the data it collects not only to make recommendations to users, but also to advise sellers on how much stock to carry, how to price goods, what goods to keep producing, and how to best market themselves.4 Amazon tracks customers and aggregates the data from all its users to see who is buying what and when. This collection of information from all the users of its site is what informs the algorithm and allows Amazon to make recommendations based on what others have purchased.5 By drawing upon the entire pool of data from all of their customers, Amazon is able to make recommendations to new customers who have yet to demarcate their preferences.

In addition to helping intensify the circulation of goods and manage stock, pricing, and other production-side concerns, target marketing serves as a technology of classifying users according to the likelihood they will provide a return on capitalist investment. Target marketing thus reinforces biases and discriminatory practices prevalent in financial institutions in order to maximize profit through the aggregation of data produced by users during seemingly innocuous acts of consumption and online attentiveness. Service providers use what Bill Davidow describes as “personal redlining” to limit choices and differentially price goods.6 Subjects who are perceived as more likely to consume are presented with better options, incentives, and prices, while others who are perceived as risks or whose browsing behavior, demographic data, and patterns of consumption do not appear as viable opportunities for creating profit are presented with different information.

The supposed value of target marketing is that it can disaggregate demand in order to determine the price a consumer is most willing to pay. User data provides a new site of information that can be mined for the purposes of forecasting demand, assigning risk, and determining prices. Differential pricing, meaning the process whereby consumers are sold the same goods for different prices, allows for the maximization of profit. For economists like Hal Varian, differential pricing is egalitarian:

Forcing a producer to sell to everyone at the same price may sound like a good idea. But it can easily end up encouraging the producer to sell only to the high end of the market. Differential pricing gives the producer an incentive to supply the product to everyone who is willing to pay the incremental cost of production . . . Forcing a policy of flat pricing in an industry where it is inappropriate due to the nature of technology may well have perverse consequences.7

Varian’s argument is that differential pricing allows the majority consumers to enjoy the same goods by correlating the price of a product to the consumer’s means. This model of differential pricing is dependent upon the collection of data about consumers in order to determine the highest price each consumer would be willing to pay. Hal Varian’s egalitarianism neglects to mention the common practice of redlining in industries such as insurance, health care, and banking, where subjects are denied access to services because they are perceived as financial risks.8 Rather than overt discrimination, “companies can smuggle proxies for race, sex, indebtedness, and so on into big-data sets and then draw correlations and conclusions that have discriminatory effects” using third-party data sources concerning buying history to predict health status, for instance, that then affects insurance rates.9 In this sense, differential pricing can be situated within a wider set of practices that use information about consumers to manage the options and choices they are presented with, particularly to incentivize those most likely to provide a return on capitalist investment, and to manage the kinds of services and options (or lack thereof) offered to those deemed too risky.

Despite the constant refrain that algorithmic models are neutral and objective, they are, as Frank Pasquale argues, “predictably biased toward reinforcing certain hierarchies of wealth and attention.”10 The outcomes of algorithmic sorting of information and market choices function as a microcosm of larger structural inequalities. But by anonymizing and aggregating data to formulate predictive models of user behavior, software platforms are able to argue that they uphold the legal protections afforded to users in regard to individual privacy. Subjects are anonymized and fragmented according to their informational byproducts within communication networks, and this makes possible the regulation of choice and potential increase in profit in ways that are not illegally discriminatory. However, the ability to differentially price goods and determine which users are most likely to provide a return on capitalist investment necessarily entails the restriction of choice and opportunity. Additionally, algorithms are predicated on deriving profit from an aggregate of consumers who are anonymized to the extent that identifying information is reconstituted into abstract data. Companies are able to claim as their private property the detailed information about their consumer base and then use it for managing the circulation of capital.

Target Marketing As Control Technology

It would be worthwhile to briefly historicize target marketing as a technique of control emerging out of the need to manage markets. Beginning with nineteenth-century industrial production, it was necessary to develop bureaucratic control over information to make production and market expansion more efficient. Max Weber famously detailed this phenomenon when he described the emergence of highly rationalized bureaucracies in reaction to the growth of industrial societies and the need for control over information in order to administer them.11 After World War II, control began to shift from conventional models of bureaucratic organization—highly rationalized organizations governed by supervision and impersonal rules for workplace conduct and information management—to computer technology.12 Beniger situates the rise of the computer and information processing technologies as tools to manage crises in the production and distribution of goods as the market’s scope of distribution spread. For Beniger, the development of mass communications technologies was essential for stimulating and reinforcing demand for mass-produced goods.

Twentieth-century mass feedback technologies, including questionnaires, house-to-house interviewing, opinion surveys, and other technologies for monitoring consumer behavior, were a further manifestation of bureaucratic rationality—a form of administration and control based on logical and statistical approaches to human behavior. The prevailing attitude concerning consumer behavior also shifted in the twentieth century from the assumption that consumers act based on reason (the self-interested rational utility-maximizer initially proposed by Adam Smith and Jeremy Bentham) whose actions could therefore be easily predicted, to the idea that consumers acted based on uncertainty.13 This necessitated, according to Beniger,

continuous monitoring to detect habituation to messages and other changes in preferences and habit. Just such monitoring of mass populations had begun to develop by the turn of the century in what would become the most widely used of all market feedback technologies: survey research.14

With target marketing, meaning the use of mass online surveillance to deliver content, goods, services, and advertisements to target markets, companies no longer need to invest in consumer polling and survey research because the consumer produces data about her preferences when she consumes, either directly through purchases or through her clicks, likes, cursor hovers, and browsing behavior.15

Whereas prior to consumer-centered information technologies, manufacturers and suppliers had the best information about demand, sales, and competition, the automatic collection of consumer data shifted power to the retailers that control these information flows, allowing them to demand greater flexibility from manufacturers and control labor costs.16 By the 1990s, the expression “data mining” became popularized in mainstream culture, and by 2005 companies would begin competing using extensive analytics and algorithms to mine data and produce valuable information for managing warehouses, transportation infrastructure, and industrial rhythms.17

Under conditions of post-Fordism, information aggregation provides greater flexibility and specialization. Capitalism shifts its focus in the overdeveloped world from production to circulation, meaning the ability to manage the speed and efficiency of the distribution of commodities, and with target marketing, we can add the allocation of advertisements and market choices. According to Jasper Bernes, the circulationist production philosophy “aims to submit all production to the condition of circulation, pushing its velocity as far toward the light-speed of information transmission as possible.”18 It becomes capital’s strategy, as Marx foreshadows in the Grundrisse, to strive

simultaneously for a greater extension of the market and for greater annihilation of space by time . . . There appears here the universalizing tendency of capital, which distinguishes it from all previous stages of production. Although limited by its very nature, it strives towards the universal development of the forces of production, and thus becomes the presupposition of a new mode of production, which is founded not on the development of the forces of production for the purpose of reproducing or at most expanding a given condition, but where free, unobstructed, progressive and universal development of the forces of production is itself the presupposition of society and hence of its reproduction; where advance beyond the point of departure is the only presupposition.19

The emphasis in post-Fordism on circulation is part of how capital restructures its organization of production so as to gain greater profit from other points in the circulation of capital. User data enables capitalists to more effectively allocate goods, services, and advertisements, and to engineer market choices so as to most likely provide a return on capitalist investment. Data mining is an essential part of what Jonathan Beller describes as the “financialization of culture,” where “attention, interiority, self-image, imagination, social practices, relationships, and time” produce value through the collection of data about the subject that is used to segment and manage the market.20

Target marketing can be framed as a technology of risk management, where risk, as Randy Martin defines it, “can be distinguished from uncertainty as an expected outcome whose likelihood or value can be quantified. For risks to be reliably calculable, the future must look like the present.”21 Target marketing is a technique of turning the uncertainty about whether past consumer behaviors are likely to be reproduced in the future into risk that can be quantified, analyzed, and hedged. Beller also argues that advertising can be considered an “instrument of risk management” given that “risk management techniques account for the vagaries of subjective actors and intersubjective social dynamics by creating a spread. They are price indexes of volatility, calculi of capture networked via screens.”22 Target marketing helps capitalists to predict user behavior and modify the distributions of options and choices through the comparative analysis of user data. Target marketing also brings the future into the present so as to manage it by modulating the future choices of consumers using predictive analytics of user behavior.

If logistics can be traced to capital’s desire to expand the market and annihilate space through time, this necessitates new means of communication and control over the distribution and allocation of goods, services, and advertisements. Logistical management is, as Bernes argues, “fundamentally different than other ensembles such as the Fordist factory; it saves on labour costs by decreasing the wage, rather than increasing the productivity of labour.”23 In the case of user data—particularly when data is used for both target marketing and managing supply chains and sites of distribution—wages are indeed decreased through logistics in the sense that capitalists are better able to forecast the amount of workers necessary to fulfill anticipated demand, thus cutting down on labor costs. However, data can also be used to intensify productivity expectations on workers.

Amazon is one example of a company that relies upon target marketing in order to accumulate profit. Amazon uses the data that users produce about themselves, particularly which goods are most frequently bought together and how user purchasing habits compare, in order to distribute recommendations and to minimize the inefficiencies of their warehouses. Amazon’s ability to offer lower prices and to increase capital gains is contingent upon the maximum productivity of their warehouses, made possible through the constant aggregation of data. The production of surplus value within the factory is accompanied by the production of information assets during both waged time and non-waged leisure time. But, as I will argue below, user data is better understood through the framework of logistical management rather than as a form of unremunerated user labor in the digital economy.

Attention Theories of Labor

One of the first scholars to address the idea that audiences produce value for capitalism is Dallas Smythe. Smythe identified what he called a “blindspot” in Marxist theory given that mass media communications were being analyzed as merely ideological or superstructural rather than through an historical materialist lens.24 For Smythe, the audience was produced as a commodity that could be delivered and sold to advertisers, and thus he conceptualized watching time as a form of work under capitalism.25 Smythe’s analysis helped to incite a debate regarding the applicability of the Marxist labor theory of value for explaining the value produced out of audience attentiveness, and this debate has been revitalized in response to the rise of online advertising.

Christian Fuchs conceptualizes looking as a form of labor in the context of leisure time online, arguing that, “if the commodity of the mentioned Internet platforms is user data, then the process of creating these data must be considered to be value-generating labour.”26 In Fuchs’ conceptualization of consumer attention as labor, the measurement of looking is a form of labor on behalf of the subjects paying attention. When Fuchs considers the relationship between what he describes as digital labor and time, he argues that all time is both the reproduction of labor power and labor time, given that data commodities are produced by social media at all times:

On Facebook and Twitter, the consumption process of the service entails all online communication and usage time. All of this time is not only reproduction time (i.e. time for the reproduction of labour-power), but at the same time labour time that produces data commodities that are offered by Facebook and Twitter for sale to advertising clients. In the consumption process, the users do not just reproduce their labour-power but produce commodities. So on Facebook, Youtube, and Twitter, all consumption time is commodity-production time.27

Fuchs goes on to say that all time on social media therefore constitutes work time, given that all time is put in the service of profit realization and the production of data commodities, and that this necessarily entails constant surveillance.28

Beller also posits that labor can be used to describe the work of attention that produces value for capital. Beller’s attention theory of value finds “in the notion of ‘labor,’ elaborated in Marx’s labor theory of value, the prototype of the newest source of value production under capitalism today: value-producing human attention.”29 For Beller, the new frontier of capital is the commodification of the human body’s capacity for attentiveness. In order to explain this point, Beller expands Marx’s notion of the labor theory of value to include the commodification of attention. Beller’s argument concerns the technologies not only of cinema but also of television, radio, computers, and the Internet, which for him are the “deterritorialized factories in which spectators work, that is, in which we perform value-productive labor.”30

Target marketing commodifies attention in that the informational byproduct from one’s online attentiveness is a source of data that can then be packaged, sorted, sold, and used to help allocate goods, services, and commodities. But in contrast to Beller’s understanding of attention as value-productive labor, where looking is a form of labor because it is productive of capital, I argue that in the case of target marketing, capital’s ability to put leisure time outside the wage relation to use through information aggregation is best understood as part of an expansion of rentier capitalism, where data is under the monopoly control of a given platform or retailer and leased out.  As David Harvey explains, “Monopoly rent arises because social actors can realize an enhanced income stream over an extended time by virtue of their exclusive control over some directly or indirectly tradable item which is in some crucial aspects unique and non-replicable.”31 Platforms have proprietary ownership over the data they collect on their users that they can then use to leverage over advertisers and suppliers alike.

As an alternative to audience labor theories of value popularized by Smythe, Beller, and Fuchs, where attention is conceptualized as a form of labor, Jakob Rigi and Robert Prey propose that:

The money paid by advertisers to media is perhaps best understood as an exchange of rent for hope: the potential of generating greater future sales. Instead of the audience being the commodity, we argue that advertising space (in the case of press media) or advertising time (in the case of television) is the commodity. The price of such advertising space or time is dependent on the projected profile of the readers/viewers attracted to this space/time. Class, gender, generation, race, national differences, and corresponding cultural habituses, among other factors, are all major aspects of audiences’ profiles.32

Given that there is a lack of any correspondence between the price of ads and the time spent online, Rigi and Prey argue that the labor theory of value is inapplicable for describing the unremunerated activity of users online. In order for the labor theory of value to apply, the activity of the audience has to produce value, and thus there must be a quantifiable measurement of time that corresponds to the time spent viewing. As Marx writes, “How, then, is the magnitude of this value to be measured? By means of the quantity of the value-forming substance, the labour, which it contains. This quantity is measured by its duration, and the labour-time is itself measured on the particular scale of hours, days, etc.”33 Labor and time are therefore inextricably linked, for Marx. But in the context of target marketing, the time spent viewing does not impact the value of user data; there is no relation between the value of advertising space/time as a commodity and the time spent online. The price of an ad is thus “a rent paid for advertising space/time, the magnitude of which primarily depends on the sociocultural profile of the audience . . . such data is best understood as a rent extracted through various mechanisms of monopoly.”34 Given that there is no temporal measurement used to assess attentiveness online, profit from audience data is not produced out of watching but out of the ability to gain rents in exchange for access to the data itself.

With target marketing, companies can raise advertising rates, i.e. extract more rent, if companies believe their ads are targeted to consumers most likely to provide a return on capitalist investment. As Chih-hsien Chen explains,

the main purpose of advertising expenditure is to prevent a realization crisis…Like speculative businesses, commercial media systems provide outlets for uncommitted capital – not as the passive absorption of surplus, but as the active speculation for future exploitation.35

Platforms thus exercise monopoly ownership over the data that they aggregate from their user bases and accumulate rent from advertisers and financiers in exchange for access. Additionally, target marketing allows for capitalists to save on the unproductive labor costs of advertisers, meaning labor which

in the process of pure circulation does not produce use-values, therefore cannot add value or surplus value. It does not add to the production of use-values because it arises specifically with commodity production out of the problems of realizing the value of commodities.36

Given that target advertising automates much of the data collection, capitalists no longer need to invest in consumer polling and survey research and thus reduces overhead. Additionally, successful advertising helps to speed up the circulation of commodities by successfully pairing consumers with goods and services they are more likely to purchase. Rendering subjective behavior more predictable increases the possibility for economic returns.

For Mark Andrejevic, this aggregation of data can be described as a process of digital enclosure, “whereby activities formerly carried out beyond the monitoring capacity of the Internet are enfolded into its virtual space.”37 It is the creation of “an interactive realm wherein every action and transaction generates information about itself.”38 To access the digital enclosure as a user is to be willingly or unknowingly subject to conditions of surveillance. The digital enclosure, Andrejevic argues, is meant to gesture to the land enclosure movement that marks the transition from feudalism to capitalism, “the process whereby over time communal land was subjected to private control, allowing private landowners to set the conditions for its use.”39 There is a parallel between the ways that land was enclosed so that revenue could be extracted through the leasing of this land as rent, and that of digital enclosure, where data is treated as the private property of software platforms and corporations, and may be leased to advertisers and financial service providers.  But it is not audience activity that produces value, but the construction of the “audience image” of an idealized, segmented audience providing anticipated returns that advertisers promise and profit from.40

The rent framework for understanding the political economy of social media thus prevents the conflation of novel forms of consumer activity related to data aggregation with labor. The understanding of user-generated data as a product of user labor, I argue, can flatten out distinctions between the capitalist structuring of labor and leisure and its relationship to time. Conceiving of consumer activity as labor would require a wholesale rethinking of the categories of value, labor, and capital. Rather than seeing target marketing as a technique for exploiting the labor of looking, I argue that in many cases it operates as a laborsaving technology once it is conceptualized as a technique of logistical management.

Target Marketing As A Laborsaving Technology

Thus far, this article has sought to conceptualize why user data is better understood within the framework of rent as opposed to labor. It is also the case that user data has downstream effects on actual sites of labor. For instance, the use of consumer data for arbitraging the labor market is evident in Mac McClelland’s account of her time as a fulfillment center worker at Amalgamated Inc., a third-party warehouse partnered with online retailers. These retailers use the data that users produce about themselves when consuming and browsing online not only to know how to best pre-position goods through user preferences and through the monitoring of which items are most frequently bought together, but also to minimize the inefficiencies of their warehouses. There is, according to Mac McClelland, a journalist who infiltrated Amalgamated Inc., no room for inefficiencies because the ability to offer lower prices and to increase capital gains is contingent upon the maximum productivity of their warehouses, which is made possible through the constant aggregation of data from every point in the circulation of capital. Consumer data analysis allows retailers to “track products and reduce operational cost while also serving as a tool for product promotions through various digital platforms.”41 Through the data collected about user and worker behavior, firms are able to determine “the exact number of humans it should take to fill this week’s orders if we work at top capacity.”42 Retailers are thus able to arbitrage the labor market by relying on temp agencies that use consumer demand trends to determine the fewest number of employees needed in order to maximize profits. As McClelland explains,

Maximizing profits means making sure no employee has a slow day, means having only as many employees as are necessary to get the job done, the number of which can be determined and ordered from a huge pool of on-demand labor literally by the day. Often, temp workers have to call in before shifts to see if they’ll get work. Sometimes, they’re paid piece rate, according to the number of units they fill or unload or move. Always, they can be let go in an instance, and replaced just as quickly.43

Companies like Amazon are able to offer free shipping, speedy delivery, and low prices to consumers precisely because of their ability to cut down on labor costs by relying on temporary, precarious, part-time work that responds reflexively to levels of consumer demand.

While both workers and consumers are subject to technologies of surveillance, consumers are nudged—meaning guided by structures of incentives—through the choice-making architecture of target marketing.44 Workplace surveillance, on the other hand, disciplines workers through the relationship between wages and the quantification of labor-power through indices of time:

Lunch is not 30 minutes and 1 second—that’s a penalty-point-earning offense—and that includes the time to get through the metal detectors and use the disgustingly overcrowded bathroom—the suggestion board hosts several pleas that someone do something about that smell—and time to stand in line to clock out and back in.45

Workers in fulfillment centers are continuously tracked not only through the use of punch cards for clocking in and out, but through the use of scanners that determine how long workers take to move and pack various products:

Dallas sector, section yellow, row H34, bin 22, level D: wearable blanket. Battery-operated flour sifter. Twenty seconds. I count how many steps it takes me to speed-walk to my destination: 20. At 5-foot-9, I’ve got a decently long stride, and I only cover the 20 steps and locate the exact shelving unit in the allotted time if I don’t hesitate for one second or get lost or take a drink of water before heading in the right direction as fast as I can walk or even occasionally jog.46

The data from the scanners is also used to determine productivity goals that seem to constantly intensify, and that require the worker to internalize this form of clock discipline by moving as fast as possible and maximizing the efficiency of all time spent “on” and “off” the clock. Additionally, as determined by the recent, and unanimous, 2014 Supreme Court decision, businesses like fulfillment centers do not have to compensate employees for the time spent waiting in line to enter and exit the workplace as it was thought not to be “integral and indispensible” to the workers’ jobs, despite being a mandatory part of the workday.47 Similarly, in the case of the class action lawsuit against Apple for unpaid time during bag searches, it was determined that since workers have the “choice” not to bring a bag to work, Apple is not responsible to pay workers for their time being searched.48 The use of information aggregation both inside and outside the workplace demonstrates how societies of control are concerned with the rapid and flexible accumulation of information during leisure time in order to manage consumer desire as well as extract surplus value created by physical labor.49 But the discipline of the clock is what differentiates labor-time from leisure-time surveillance.

Viewing leisure-time data aggregation as a form of logistical management—rather than unremunerated labor—clarifies the distinction between the commercial surveillance of leisure time and workplace surveillance of labor structured by the wage relation.  Both forms of surveillance contribute to the circulation of capital but under conditions that are radically different. McClelland clearly conveys the ways that workers are subject to conditions of surveillance that correlate to the fact that their time is not conceptualized as “free” but owned by capitalists, and therefore highly regimented, quantified, and enforced in order to ensure productivity gains and “satisfactory” customer experiences.50 In contrast, online users browsing the web during leisure time are not subject to the same time constraints and conditions, given that this time is conceptualized as “free” leisure time. Thus, while workers are disciplined through surveillance that enforces the equivalence of the wage with time and have little to no control over their working environments, consumers are managed through their ability to make choices. The data accumulated about consumer demand provides a means of intensifying the work-place regulation of workers like McClelland through ever-increasing productivity goals and makes possible the predetermination of the amount of temporary workers necessary while simultaneously working to distribute risks and rewards to consumers based on their data profiles.


This analysis of target marketing and its dependency on leisure time surveillance illustrates the ways ordinary life is increasingly subjected to technologies of surveillance. Ubiquitous surveillance is a key characteristic of post-Fordist societies of control, societies organized according to the flexible accumulation of capital made possible by information technologies like target marketing. Target marketing makes subjective behavior classifiable, marketable, traceable, and legible, ultimately containing, managing, and exploiting the productive power of subjects while concentrating power over suppliers, laborers, and consumers alike through logistical management. I would like to close with a brief discussion of the political implications of an analysis that treats user activity as a form of unremunerated labor.

Online users in Fuch’s model are victims of infinite exploitation because they produce unremunerated user generated content and produce information that is sold as a commodity to third-party advertisers:

While no product is sold to the users, the users themselves are sold as a commodity to advertisers. The more users a platform has, the higher the advertising rates can be charged. The productive labor time that is exploited by capital, on the one hand, involves the labor time of the paid employees and, on the other hand, all of the time that is spent online by the users.51

This reading seems to contradict the idea that what scientific knowledge and technology provide is precisely not the lengthening of the hours of work (and therefore absolute surplus value) but rather raise the productivity of already existing labor.52 This model of user labor also results in the understanding that users are subject to a rate of exploitation that “converges toward infinity” given that users “work completely for free.”53 Fuch’s understanding of exploitation makes it difficult to distinguish the different conditions of labor and leisure that structure relations of power in the digital economy and suggests that one hundred percent of an online user’s time is hyper-exploited surplus labor time.

However, one should hesitate to fold the data trails, meaning the residues of practices of online consumption and attentiveness, into a form of labor under generalized conditions of capital accumulation. In contrast to the argument that labor can be used to describe the value produced out of user-data online, the intervention this paper makes is to consider user attention as part of a logistically coordinated digital economy in which the profit accumulated through user activity online is better understood as a labor-saving technology to the extent that it can be used to regiment the allocation of advertisements, goods, and services, and allows for the increased rationalization of labor within fulfillment centers, supply chains, and advertising. This understanding of the digital economy works to refocus attention on the connections between the advertising industry, fulfillment center labor in the U.S., and the extraction of data from the entire network of workers and consumers rather than emphasizing a generalization of labor, which works to flatten these distinctions. I hope that by situating the extraction of profit from user attentiveness and online consumption within the circulation of capital, this framework can attend to the qualitative and quantitative distinctions between sites of production, distribution, and consumption, and their mutually reinforcing logics and technologies of surveillance in the logistically coordinated world of target markets.


This article received comments during the peer review process and from the editors of Lateral that were instrumental to the piece’s development. The article is also indebted to the intellectual guidance of Professors Robert Meister, Carla Freccero, Warren Sack, Mark Andrejevic, and Surya Parekh.


  1. Anna Tsing. “Supply Chains and the Human Condition,” Rethinking Marxism: A Journal of Economics, Culture, and Society 21, no. 2 (2009): 149.
  2. Ibid .
  3. Deborah Cowen, The Deadly Life of Logistics: Mapping Violence in Global Trade (Minneapolis: University of Minnesota Press, 2014), 2.
  4. Leena Rao, “How Amazon is Tackling Personalization and Curation for Sellers on its Marketplace,” Tech Crunch, Aug. 21, 2013,
  5. Ibid.
  6. Bill Davidow. “Redlining for the 21st Century,” The Atlantic, March 3, 2014,
  7. Hal Varian, “Differential Pricing and Efficiency.” First Monday 1, no. 2 (1996),
  8. Brynjolfsson, Erik and Brian Kahin, Understanding the Digital Economy: Data, Tools, and Research (Cambridge, MA: MIT Press, 2002).
  9. Astra Taylor and Jathan Sadowski, “How Companies Turn Your Facebook Activity into a Credit Score” The Nation, May 27, 2015,
  10. Frank Pasquale, The Black Box Society: The Secret Algorithms That Control Money and Information (Cambridge, MA: Harvard University Press, 2015), 188.
  11. Max Weber, Economy and Society (Berkeley: University of California Press, 1978).
  12. James R. Beniger, The Control Revolution: Technological and Economic Origins of the Information Society (Cambridge, MA: Harvard UP, 1986), 7.
  13. Herbert A. Simon, “Rational Choice and the Structure of the Environment,” Psychological Review 63, no. 2 (1956): 129–138.
  14. Beniger, The Control Revolution, 386.
  15. Mark Andrejevic, “Watching Television Without Pity: The Productivity of Online Fans,” Television New Media 9, no. 1(2008): 24–46.
  16. Nelson Lichtenstein, The Retail Revolution: How Wal-Mart Created a Brave New World of Business (New York: Metropolitan, 2009), 57.
  17. Gil Press, “A Very Short History of Data Science,” What’s The Big Data, April 26, 2012,
  18. Jasper Bernes, “Logistics, Counterlogistics and the Communist Prospect,” Endnotes, September 2013,
  19. Karl Marx, Grundrisse: Foundations of the Critique of Political Economy (New York: Vintage, 1973), 540.
  20. Jonathan Beller, “Informatic Labor in the Age of Computational Capital,” Lateral 5, no. 1 (2016),
  21. Randy Martin, An Empire of Indifference: American War and the Financial Logic of Risk Management (Durham: Duke University Press, 2007), 4.
  22. Beller, “Informatic Labor.”
  23. Bernes, “Logistics, Counterlogistics.”
  24. Dallas Smythe, “Communications: Blindspot of Western Marxism,” Canadian Journal of Political and Social Theory 1 (1977): 1.
  25. Smythe, “Communications,” 3.
  26. Christian Fuchs, Digital Labor and Karl Marx (New York: Routledge. 2014), 246.
  27. Ibid., 90.
  28. Ibid., 101.
  29. Jonathan Beller, The Cinematic Mode of Production: Attention Economy and the Society of the Spectacle (Hanover, NH: Datmouth College Press, 2006), 5.
  30. Ibid., 1.
  31. David Harvey, “The Art of Rent: Globalization, Monopoly and the Commodification of Culture,” Social Register 38(2002): 94.
  32. Jakob Rigi and Robert Prey, “Value, Rent, and the Political Economy of Social Media,” The Information Society 31, no. 5 (2015): 396.
  33. Karl Marx, Capital, Vol. I (New York: Random House, 1977), 129.
  34. Rigi and Prey, “Value, Rent, and the Political Economy,” 392.
  35. Chih-hsien Chen, “Is the Audience Really Commodity? An Overdetermined Perspective of the Television Economy” (Paper presented at the Annual Meeting of the International Communication Association, Marriot Hotel, San Diego, CA, 2003).
  36. Ian Gough, “Marx’s Theory of Productive and Unproductive Labor,” New Left Review 1, no. 76 (1972): 60.
  37. Mark Andrejevic, Reality TV: The Work of Being Watched (Lanham, MD: Rowman & Littlefield, 2004), 238.
  38. Mark Andrejevic, ISpy: Surveillance and Power in the Interactive Era (University Press of Kansas: KA, 2007), 2.
  39. Ibid., 3.
  40. Chen, “Is the Audience Really a Commodity?”
  41. John Langley, Jr., Third-Party Logistics Study: The State of Logistics Outsourcing (2015), 21, accessed March 28, 2017,
  42. Mac McClelland, “I Was A Warehouse Wage Slave,” Mother Jones, March/April 2012,
  43. Ibid.
  44. See Cass Sustein, “Nudging: A Very Short Guide.” Consumer Pol’y 583 (2014): 1–7.
  45. McClelland, “I Was A Warehouse Wage Slave.”
  46. Ibid.
  47. Alison Griswold, “Supreme Court Decides Amazon Workers Don’t Need to be Paid While Waiting for Mandatory Security Screenings, Slate, December 9, 2014,
  48. Reuters, “Apple Defeats Class Action Lawsuit Over Bag Searches,” Fortune Magazine, November 8, 2015,
  49. Grant Vetter, The Architecture of Control: A Contribution to the Critique of the Science of Apparatuses (Winchester, UK: Zero Books, 2012).
  50. McClelland, “I Was a Warehouse Wage Slave.”
  51. Christian Fuchs, “Labor in Informational Capitalism and on the Internet,” The Information Society 26 (2010): 191.
  52. Karl Marx, Grundrisse, 708-709.
  53. Christian Fuchs, “Labor in Informational Capitalism,” 191.
Lindsay Weinberg

Lindsay Weinberg

Lindsay Weinberg is a fourth year PhD Candidate in the History of Consciousness Department at the University of California Santa Cruz. Her research concerns how personalization—the web of technologies, cultural practices, and economic logics that generate information about consumers to market goods and services to targeted audiences—is part of a larger cultural and economic transformation under capitalism. She is the Contributing Editor of Studies In Control Societies.