Digitalized Markets

Johan Hagberg, Hans Kjellberg, Johan Hagberg, Hans Kjellberg

• Publisher: Routledge
• Year: 2021
• Print ISBN: 9780367655716
• eBook ISBN: 9781000351392
• Topic: Business
• Subtopic: Business generale
• Index: Business

This book addresses how digitalization influences markets, and attempts to put research on digitalized markets center-stage. It explores digitalized markets through empirically based theorizing concerning the consequences of digitalization for mundane markets. The individual chapters explore several mundane markets, including personal transportation, temporary accommodation, fashion clothing, concert tickets, and web shopping. They employ a variety of useful concepts and methods to approach the complexity of digitalization of markets.

Based on these accounts, the digitalization of markets is conceived as comprising transformation of three main aspects of markets. First, digitalization transforms the elements of markets, such as actors, devices, objects, and places that contribute to constitute markets. Second, digitalization alters market processes, or developmental event sequences by changing the activities that contribute to produce the market and thus how markets develop and take form. Third, digitalization has implications for the overall forms that markets assume in terms of how market elements and processes are linked and organized.

The volume provides important contributions to our understanding of digitalized markets both through rich empirical accounts of a variety of market contexts and through conceptual developments that improve our ability to analytically deal with the market consequences of digitalization.

The chapters in this book were originally published as a special issue of Consumption Markets & Culture.

Cookie monsters. Anatomy of a digital market infrastructure

Kevin Mellet and Thomas Beauvisage

ABSTRACT
The aim of this article is to examine the market arrangements built by the online marketing industry around small pieces of data now ubiquitous in digital markets–namely “http cookies.” We show how cookies have become the backbone and the main vehicle of a vast market infrastructure, based on its ability to transform online behavioral information into data assets, and to attach these assets to advertising products. We examine the complex trading operations that are implemented from the elementary brick that constitutes the cookie. We also raise the question of the strength and durability of this infrastructure, at a time when it is disputed and seems weakened. Beyond the particular case of cookies, we identify three main operations that market infrastructures typically support: knowledge production, capitalization, and coordination. We also highlight the centrality of “datafication” (tracking, “data lake” building, matching, etc.) in the process of market digitalization. We thus contribute to the framing of the concept of (digital) market infrastructure.

1. Introduction

We are talking about an invention of the 1990s that the world of advertising has appropriated and which it has diverted from its original purpose. A Frankenstein monster, not really effective, but on which we have built so much value that it has always been difficult to reform it. An advertising executive, about the cookie¹

Http cookies, or simply cookies, are small pieces of information placed and stored on web browsers by websites as users browse pages. Since their invention in 1994 by a Netscape developer, they have had a remarkable fortune: their use is almost universal in areas such as web session management, audience measurement or content customization (Schwartz 2001; Greenstein 2015). But among all the application domains of cookies, online advertising is certainly the industry that has designed the largest and the most sophisticated cookie-dependent architecture (Turow 2012; Smith 2014). Over the past twenty years, this industry has come to rely more and more on cookies, to the point of becoming totally addicted to them²: the commercialization of advertising inventory (especially in the sub-field of display advertising³), and the coordination between the multiple intermediaries that make up the online advertising food chain, is very much based on the mobilization and circulation of cookie-based information (Englehardt and Narayanan 2016). In addition, some so called “third-party” players have managed to form relationships with large networks of websites, allowing them to build databases recording information on virtually any Internet user (Falahrastegar et al. 2014; Binns et al. 2018). It is common that the mere opening of a page on a website generates the production or updating of several dozen cookies. And advertising market players themselves come to observe that cookies proliferate, and that this proliferation might be a problem in itself (IAB 2014).

This article describes the process of “cookification” of the online advertising industry. Following a science and technology studies (STS) perspective, we show how this innocuous piece of code has become the backbone and the main vehicle of a vast market infrastructure, based on its ability to transform online behavioral information into data assets, and to attach these assets to advertising products. We examine the complex trading operations that are implemented from the elementary brick that constitutes the cookie. Finally, we raise the question of the strength and durability of this infrastructure, at a time when it is disputed and seems weakened by the growth of mobile uses and the establishment of new regulations. Beyond the particular case of cookies and the advertising industry, we show the interest to pay particular attention to market infrastructures, basically defined as the material and discursive objects and assemblages that create the grounds on which markets silently operate. Following recent conceptualizations by Chakrabarti et al. (2016) and Kjellberg, Hagberg, and Cochoy (2019), this article seeks to answer the following two questions: What does the particular case of cookies bring to the general understanding of market infrastructures? What lessons can we draw from the point of view of the study of the digitalization of markets?

Our argument is that the history of the “cookification” of the online advertising industry illustrates key properties of digital market infrastructures. In many ways, “cookification” appears to be a very traditional process of infrastructuring, the success of which is measured by its ability to enroll new players, to absorb old devices and mechanisms, and to foster new uses (Star and Bowker 2002). Today, there is no business, online or brick and mortar, which escapes the influence of cookies. The ubiquity, invisibility (in the eyes of ordinary Internet users, at least), and permanence of cookies contribute to its status as a market infrastructure (Kjellberg, Hagberg, and Cochoy 2019).

Our study allows us to go further into the elaboration of this conceptual framework. In the fields of marketing and advertising, cookie databases have a special function: they allow consumers to be described from the recording, aggregation and analysis of their web browsing activity, within a web site or from site to site. Cookies thus make it possible to produce a limited but extensive knowledge of Internet users. This knowledge is stable and transferable, through time and space, because the cookie contains a unique identifier that allows whoever manages it to find a user again on another website.⁴ Through the multiple matching and aggregation operations that it allows, for example when cookies are pooled into audience segments, this identifier serves as a powerful means for operations of capitalization and valuation by the various market players involved in the online advertising food chain. Finally, cookies allow marketers and advertizers to “follow” users and to address them with various forms of targeted marketing and advertising. In other words, the cookie is at the center of a socio-technical system that feeds a knowledge base produced from the tracking of consumers, the capitalization of this knowledge into marketable segments, and its operationalization in advertising and marketing actions. We argue that the combination of these operations of knowledge production, capitalization and coordination, turn the cookie into a very powerful market infrastructure.

The article is structured as follows. Section One presents our theoretical framework and research objectives. We clarify our use of the concept of market infrastructure, which occupies a central place in our approach. Section Two outlines our data sources and methodology. The next two sections describe the “cookification” of the online advertising industry, through the history, the effects and the legacy of the cookie infrastructure. Section Three describes the formation of a cookie-based advertising market infrastructure by telling the different steps that led to the widespread adoption of cookies by advertising professionals. We analyse how the advertising industry has opportunistically succeeded in integrating the cookie in a very diverse set of activities, and thus contributed to turn it into a market infrastructure. Section Four raises the question of the solidity and sustainability of the cookified market infrastructure as it is challenged. In doing so, we seek to characterize and examine the causes of the power of cookies as a market infrastructure. Finally, we provide a summary and conclusions.

2. Theoretical framework and research objectives

To what extent can an analysis of cookies, and the ongoing process of “cookification” of online advertising, shed light on the shaping and operation of digital markets? In this section, we present our theoretical framework, focused around the concept of market infrastructure. First, we provide a brief survey of the literature on the movement of “datafication,” with a focus on works that are concerned with marketing and advertising. Then we are interested in the concept of market infrastructure. We recall the genealogy of the concept and specify how we contribute to its theoretical development.

2.1. Datafication and the marketing industry

Big data technologies, and especially those using personal data, have received growing academic attention in recent years. Many studies have investigated the social and economic consequences of digital knowledge infrastructures, tracking devices, and algorithms, using the term of “datafication.” It is possible to identify two main perspectives inside this corpus.

A first set of works adopts a wide-angle lens and observes a largely coherent, intentional and convergent dynamic, towards the setting up of an infrastructure mainly dedicated to tracking and surveillance. Schneier (2015) points out that, “in the normal course of their operations, computers continuously document what they’re doing” (15); big data systematizes the apprehension of data and traces as a valuable resource requiring attention and investment. Kitchin (2014) describes how the manufacture of data is greatly linked to their use:

databases are designed and built to hold certain kinds of data and enable certain kinds of analysis, and how they are structured has profound consequences as to what queries and analysis can be performed; their formulation conditions the work that can be done through them. (…) Data infrastructures host and link databases into a more complex sociotechnical structure. (21–23)

This sociotechnical structure is the subject of a vast body of research that shares a common description of datafication as a convergent process of constitution of a vast tracking and surveillance infrastructure (Lyon 2001; Zimmer 2008; Christl and Spiekermann 2016; Zuboff 2019). These works take a critical look at this movement and denounce the risks they pose to individual freedom and privacy. This dynamic, and the controversies and socio-economic struggles in opens, are of course central. But the focus chosen by these authors tends to leave aside the specific application fields and concrete uses of these tracking systems.

A second set of works on datafication is more focused on the multiples uses of data by businesses in the practical course of their activity, especially in the area of digital marketing. These studies elaborate a more disordered, and largely unintended narrative, which leads to the recognition of the omnipresence of data in the social and economic existence of the digital economy.

As markets are digitalized, data are a key element in the operation of markets in a variety of forms: consumer records, scoring and targeting instruments, customization engines, algorithmic pricing, vending and trading machines. Hence, data act as a versatile, yet vital, apparatus for markets, through the construction of data as a device, a valuation tool, a measure, an asset, or an infrastructure. Following this perspective, Cluley (2018) defines the process of datafication as the non-marketing activities needed to create a standard unit of data. From a case study on the implementation of a measure of the viewability of advertisements on the Internet, he describes a three steps process, “where an alignment of interests stabilizes an object of measurement which is then formalized in a standard method for measurement.”

Although data collection practices for marketing purposes are hard to investigate (see below our methodological framework), attempts to assess them, or to report on the data brokerage industry, have come to constitute a cumulative knowledge base on the sociotechnical market arrangements built upon personal data. Based on a self-ethnography, Ebeling’s work (2016) shows how ordinary life events become objects of capitalization by data brokers, private companies that manufacture and maintain databases of “digital doubles” of virtually each individual consumer. These data doubles exist without the knowledge of and away from their human alter egos, and come to haunt them, more and more frequently, like phantoms (or monsters!), in the form of hyper-targeted marketing. Data doubles are enrolled in a set of economic activities, and valued as commodities or assets (Bouk 2017; Neyland and Milyaeva 2016; Fourcade and Healy 2017), or exploited as segments and marketable groups (Turow 2012; Ariztia 2015; Cluley and Brown 2015).

This disparate body of research shows how data has become a central concern for people and organizations operating and acting on markets. Yet, these works have only paid oblique attention to the materiality of these data doubles, especially to their cookie-based nature–with the notable exception of Turow (2012) and Cluley and Brown (2015). Turow’s work is focused on the advertising food chain and its successive evolutions guided towards more personalization and targeting. Cluley and Brown investigate the development of cookie-based market devices, and focus on how such technologies change the way consumers are represented, categorized and, finally, “dividualized” for marketing operations as fine-grained ad-hoc segments and specific properties.

Finally, the http cookie is also an object of interest in computer science, with works studying the mechanisms of production and management of cookies, and technical coordination between actors having use of cookies (Park and Sandhu 2000; Mayer and Mitchell 2012; Ghosh et al. 2015). Cookies have also been of interest to lawyers and privacy scholars because their legal status as personal information appears fluctuating and uncertain (Siebecker 2002; Mills 2008). In addition to this scientific literature, there are more personalized accounts of the invention and development of cookies, which place particular emphasis on the hesitations, decisions and remorse of their inventor (see in particular Schwartz 2001).

2.2. From market devices to market infrastructures

The study of market datafication can benefit from two neighboring fields which share a common affiliation with STS, but which have undergone separate developments: infrastructure studies, and “market studies.” Infrastructure studies were especially theorized as a research agenda by Star (Star and Ruhleder 1996; Star 1999), and Bowker et al. (2009). Following a socio-historical perspective, these authors set themselves the objective of putting into the foreground what, in general, is in the background, invisible and transparent. Infrastructures are described as a vast set of objects that are necessary for most human activities: material collective equipment, such as bridges, energy or communication networks, but also inscriptions such as protocols, standards and classifications. There is now a vast literature in infrastructure studies, from which there are several interesting observations for our purpose. Good and usable infrastructures tend to disappear, “by definition” (Bowker and Star 1999), which makes them particularly hard to see. Infrastructures are often in the making and expanding, and they rarely result in their actual form of a defined and centralized plan. Standards and classification are of crucial importance and they embed social and political arrangements and compromises.

The other approach has developed from the work of Callon (1998) in economic sociology and then expanded to related disciplines suc...

Table of contents

• Cover
• Title Page
• Copyright Page
• Table of Contents
• Citation Information
• Notes on Contributors
• Introduction: Digitalized markets
• 1 Cookie monsters. Anatomy of a digital market infrastructure
• 2 Conceptual controversies at the boundaries between markets: the case of ridesharing
• 3 Digitalization and the hybridization of markets and circuits in Airbnb
• 4 Sold out? Reconfiguring consumer demand through the secondary digital ticket market
• 5 (De-)stabilizing the digitized fashion market on Instagram–dynamics of visual performative assemblages
• Index