Author’s accepted manuscript. Accepted for publication in Journal of Retailing and Consumer Services. Reference: Soutjis, B., Cochoy, F., & Hagberg, J. (2017). An ethnography of Electronic Shelf Labels: The resisted digitalization of prices in contemporary supermarkets. Journal of Retailing and Consumer Services, 39, 296-304. https://doi.org/10.1016/j.jretconser.2017.08.009 -2- An ethnography of Electronic Shelf Labels. The resisted digitization of prices in contemporary supermarkets Bastien Soutjis University Toulouse Jean-Jaurès/LISST-CERS, 5 Allée Antonio Machado, 31000 Toulouse, France, bastien.soutjis@univ-tlse2.fr Franck Cochoy University Toulouse Jean-Jaurès/LISST-CERS, 5 Allée Antonio Machado, 31000 Toulouse, France, cochoy@univ-tlse2.fr (corresponding author) Johan Hagberg University of Gothenburg, Vasagatan 1, PO Box 610, SE-405 30 Gothenburg, Sweden, johan.hagberg@gu.se Abstract Contemporary retail markets have experienced and are experiencing an important digitization shift in the form of computers and associated technologies. Among a large array of digital innovations, Electronic Shelf Labels (ESLs) deserve particular attention. ESLs, despite their long history and many benefits, have not become ubiquitous. The purpose of this paper is to account for this “resisted evolution” of digitized prices. It draws theoretically upon science and technology studies, infrastructure studies, market studies, and previous literature on price representations in retailing. It draws empirically on a combination of ethnographic and historical methods. The paper shows that ESLs do not replace paper prices, but, rather, work together with them: on one hand, they compete to represent prices with their respective features, and on the other, they co-operate in order to reinforce the visibility and attractiveness of products and promotions. Key words: market studies, electronic shelf labels, price representation, retailing, visibility. 1. Introduction Contemporary retail markets have experienced and are experiencing an important digitization shift in the form of computers and associated technologies (Hagberg et al., 2016). Important contributions to this development have included the introduction of the Universal Product Code (UPC) in the 1970s (Cochoy et al., 2016a) and the subsequent spread of checkout scanners, Point Of Sale (POS) software, computerized couponing and loyalty programs, biometric payments, RFID chips, e-catalogs, e-commerce, and, more recently, QR codes, smartphone apps, and other interactive technologies. Among this large array of innovations, Electronic Shelf Labels (ESLs) deserve particular attention. These devices present themselves as electronic price tags: they allow retailers to change prices through computerized transmitters instead of tedious handheld operations, and they offer consumers a clear and accurate form of price display. As such, these devices evidence a puzzling situation. On one hand, they pertain to one of the most classic and crucial features of self-service retailing: the display of prices; thus, it would seem that they should play a distinct and prominent role in -3- retail digitization. On the other hand, despite their relevance and decades-long history, ESLs have not become ubiquitous. Even in those few retail outlets where they have been implemented, ESLs seem to face resistance from paper price tags and price cards. The purpose of this paper is to account for this “resisting evolution” of digitized prices and to discuss the role of price display technologies in the functioning of contemporary economy. Retailing has been subject to a great number of pricing innovations that have received considerable attention (see e.g. Grewal et al., 2010; 2011; Kopalle et al., 2009). The literature has explored the effects of low price guarantees (Biswas et al., 2006), dynamic pricing models (Hall et al., 2010), and mobile payments on consumers’ price images of retail stores (Falk et al., 2016), among others. However, while prior studies have examined different retail price innovations, they have so far tended to neglect the more material aspects of pricing innovations: that is, the particular devices involved in price representation. Further, there has so far been a paucity of studies addressing how digital forms of pricing interrelate with their more analogue counterparts. Moreover, so far, existing studies have paid insufficient attention to the process of innovation, including how and why particular digital technologies face resistance. Theoretically, we draw upon infrastructure studies (Bowker et al., 2010) and market studies (Callon, 1998), two fields of research inspired by science and technology and the Actor- Network Theory (ANT). More specifically, we rely on the previous literature focused on price representation practices in retailing (e.g. Grandclément, 2008; Hagberg and Kjellberg, 2015). These works portray markets not as abstract platforms connecting supply and demand, but as complex “agencements” (Callon et al., 2007; Callon, 2016) or arrangements of human actors, organizational schemes, and material artifacts that shape and perform the economic game. Within such configurations, “price representations,” price infrastructures, and price display techniques—rather than abstract and automatic “price adjustment” mechanisms—play major roles. Empirically, this paper draws on several different sources. First, it builds a longitudinal “archaeology of price tags” based on a systematic reading of the trade journal The Progressive Grocer from 1922 to the present. Second, it references an ethnographic study of three stores equipped with ESLs from one specific provider, including observations within these stores documented through field notes, photos, and interviews with retail managers and ESL professionals, as well as observations of the websites of ESL providers. Third, it is based on a short ethnography of paper price tags in New York City performed in 2016. The paper shows how ESLs prolong but also challenge nearly a century of price display technologies by improving the chain of “price representations” (Hagberg and Kjellberg, 2015) by adding new possibilities in terms of price flexibility. Further, it explores how the claims of ESL providers and the promises of their technologies should be questioned as soon as these claims and devices are put into practice, particularly when ESL faces competition from classic paper price display techniques. The paper is structured as follows. In the next section, we introduce our theoretical framework. This is followed by a section describing our method. Thereafter, our findings are presented in four sections. The first of these sections traces the history of price tags and the role of ESLs within this history, then discusses ESL providers and how and why the stores under study have been equipped with ESLs. In the second section, we further explore ESLs by entering stores and by presenting the three variants of price representation observed during our visual ethnography. The third section analyses the interaction between paper and electronic tags that represents prices in food retail markets. In the fourth section, we further -4- elaborate on these findings by putting contemporary ESLs into perspective. Finally, we conclude with some reflective remarks. 2. Theoretical framework Theoretically, this paper draws on previous work on science and technology studies on markets and innovation and price representations in retail settings, presented below in two sections. 2.1 Infrastructures, innovations, and markets First, the paper draws upon studies of infrastructure (e.g. Bowker et al., 2010) and markets (e.g. Callon, 1998), two fields of research inspired by science and technology studies and Actor-Network Theory (ANT). Science and technology studies have long explored the complex interactions between human beings and material artifacts. The so-called Social Construction Of Technology (SCOT) model suggests that technologies are neither neutral nor detached from social and political stakes. ANT proposes a more symmetrical approach in which human and non-human entities combine their properties to shape social action, without any of them a priori superseding the other (Callon, 1986). Both approaches have traditionally focused primarily on the engineers and entrepreneurs who design and implement technologies (Akrich et al., 1988; Bijker and Law, 1992; Latour, 1992). More recently however, greater attention has been paid to users, both as co-innovators (Oudshhoorn and Pinch, 2003; Von Hippel, 2005) and as the targets of technological efforts (Hyysalo et al., 2016). Within this perspective, we will examine price display techniques that will help us pay close attention to not only the views of ESL providers, but also the perspectives of their clients and ESL technology itself. Infrastructure studies insist on “collective equipment necessary to human activities,” such as road and computer networks, defined as “a broad category referring to pervasive enabling resources in network form” (Bowker at al., 2010). Considering price display equipment as infrastructures illustrates that prices are neither abstract entities nor isolated numerals, as economics tends to represent them, but are instead interconnected entities whose technological design frames consumers’ cognition and behavior. This view supports the agenda of market studies, a body of work that proposes to study markets as the performative enactments of economic and managerial theories (e.g. price models or pricing strategies) (Callon, 1998) through the use of ad hoc “calculative devices” (Callon and Muniesa, 2005). Fabian Muniesa and his colleagues further refined this view by proposing that markets should be considered as specific forms of “agencements” (Muniesa et al., 2007), a French synonym for “arrangements.” By alluding to both “agent” and “agency,” this notion emphasizes the distribution of action between human agents and technical artifacts. Price display systems, as human-driven tools that nevertheless have their own agency (e.g. their own capacity to act), clearly fall within this category. All of these works support the perspective of markets not as abstract platforms connecting supply and demand, but as complex “infrastructures” and “agencements” comprising human actors, organizational schemes, and material artifacts. These infrastructures and agencements shape and perform the economic game. Within such configurations, the major role is played not by abstract and automatic “price adjustment” mechanisms, but by price display techniques. -5- 2.2 Price representations We also draw upon previous contributions relating prices and price representation practices in retailing. Traditionally, prices have been considered by mainstream economics to be neutral outcomes of exchanges, theorized under the familiar law of supply and demand. From a marketing perspective, however, prices have instead been conceived as a tool within the marketing mix used by marketers to position products in the market (Borden, 1964). Within marketing, extensive attention has been paid to retail pricing, including the use of different pricing strategies, practices, and techniques and consumer perceptions of retail prices (e.g., Ahmetoglu et al., 2014; Biswas et al., 2006; Falk et al., 2016; Grewal et al., 2010; 2011; Hall et al., 2010; Kopalle et al., 2009; Lombart et al., 2016). However, despite this extensive body of literature, there is a relative paucity of studies addressing the actual work of representing both prices and the material aspects of price representations (Hagberg and Kjellberg, 2015). This brings us closer to the notion of “practices” (Shove et al., 2012) and back to the concept of “market agencement” (Callon, 2016). According to Shove et al. (2012), practices are combinations of materials, competence, and meaning. According to this perspective, there is no such thing as autonomous people making decisions within a given external context; rather, the social world is made of practices formed through webs of infrastructures, artifacts, culture, norms, skills, and actions (Shove, 2010). Prices are both the expressions of “practices” and the production of particular “market agencements.” As practices, prices are not reducible to managerial action, but are instead the expression of the multidimensional practice of price fixing. As aspects of market agencement, prices are not the neutral outcome of the interaction between the supposedly independent blocks of “supply” and “demand,” but represent, rather, the production of a hybrid network of objects and actions designed to value and qualify goods (Callon and Muniesa, 2005). Grandclément (2008) showed that prices are ubiquitous in retail stores by studying the concrete actions of the retailers who handle prices in what she called a “jungle of paper.” She drew attention to the fact that prices are primarily material entities, requiring paper, cardboard, plastic, etc. They also tend to involve visual, material, and discursive elements and to take multiple forms. This ubiquity and plurality led Grandclément (2008) to notice that many of the prices displayed in stores do not reflect what customers actually pay (e.g. special prices may be offered for “virtual bundles,” but many consumers still buy single goods). Hence, she distinguished between “prices of calculation” and “paid prices.” Secondly, she claimed that prices are not isolated. Rather, they are intertwined with several other dimensions through various forms of spatial arrangements embedded in different relationships among retailers, providers, suppliers, and consumers. Third, Grandclément (2008) argued that prices combine numerical and qualitative dimensions, such as colors, size, and so on. Indeed, prices are traditionally associated with the cold and rational world of calculation, rather than the warm world of non-calculation. However, according to her work, this distinction cannot hold, since prices are micro-devices that blur the boundaries between the rational and passionate motives of consumers. In order to acknowledge the plurality of price forms present in retail markets, Hagberg and Kjellberg (2015) proposed the notion of price representation. They defined price representation as “the various ways in which prices are made available to market participants, that is, renderings of prices generated and disseminated by market actors” (Hagberg and Kjellberg, 2015, p.180). They also noted that these price representations and their generations are not acknowledged by the literature, since “price representations are accepted as accurate reflections of price” (Hagberg and Kjellberg, 2015, p.180). Like Grandclément (2008), the authors stressed that price representations often differ from the prices paid by the consumer. -6- Indeed, paid prices change frequently and sometimes independently of their representations, and they can also differ from one consumer to another for the same good. Thus, the relationship between prices and their representations is complex, and the ways of representing prices may impact their levels. 3. Method In order to be able to account for both the development over time and the contemporary state of ESL, this study combined different methods. The paper also draws upon three kinds of materials. First, it constructs a longitudinal “archaeology of price tags” based on a systematic reading of the trade journal The Progressive Grocer from 1922 to the present. This approach examines the aspects and evolution of all the little “market-things” that equip consumption. It accomplishes this by looking closely at the features and promises of these items as they appear in advertisements, articles, and patents (Cochoy, 2016; Hagberg, 2016; Hagberg and Normark, 2015). Like archaeologists, we base our analysis on the material forms that price representation devices and techniques have assumed. We combine this approach with a more traditional historical analysis of written articles from the magazine. Specifically, we photographed all pages with content directly or indirectly linked to price representation, coded the resulting archive using Adobe Bridge, and analyzed various thematic lines. Second, the paper draws on multi-sited ethnography (Marcus, 1995) to identify the tension between local price contingencies as local and material artifacts and the global significance of prices as a universal form of economic information. This ethnography encompasses a study of actors (retailers) and market devices (ESL and other techniques), according to an ANT- inspired view of ethnography in which “following the actors” (Latour, 2005) means tracing both human and non-human agencies and their associations (Baiocchi et al., 2013). Along these lines, we have conducted an ethnographic study of three stores equipped with ESLs from a single ESL provider (Pricer) in the city of Gothenburg in Sweden. All three stores belonging to the ICA retailer chain: two ICA Nära and one ICA Kvantum. ICA is the major Swedish grocery retail chain, and most of its stores are equipped with Pricer ESLs. Pricer is a Swedish company and the leader in the Swedish ESL market, with approximately 80 to 85% of market share (in terms of number of equipped stores). The company, founded in 1991, claims to be the global leader in “providing in-store digital shelf-edge solutions that enhance both store performance and the shopping experience.”1 According to its website, Pricer has equipped 13,500 stores in over 50 countries, representing around 110 million electronic labels. The present study includes observations within three of these stores, documented through field notes and photos, interviews with retail managers and ESL professionals conducted from 2015 to 2016, and observations of the websites of ESL providers (primarily Pricer, which produces the electronic tags used in the observed stores, but also Altierre, DigiSystem, DisplayData and Store-Electronic-System). We showed photos taken during our observations to our interviewees in order to facilitate discussions. After the interviews, we also conducted additional email exchanges in order to complement our information. We have also conducted a short ethnography of paper price tags in New York City from March to April 2016. This part of the study consisted of visits to grocery stores located in Manhattan, documented with a camera. The results complement those of the other two studies 1 http://www.pricer.com/en/Company/ (accessed 08.06.17) -7- in terms of both time and space. Thus, this part of the study provides a contemporary picture that facilitates the comparison of the historical and the Swedish case. Overall, our combination of different sources aims to present rich and versatile material on the topic of ESLs in order to analyze the extent to which they have been adopted or resisted. 4. From the history of price tags to the present promises of ESL ESL is one of the latest innovations in a long range of pricing techniques. Therefore, to support a comprehensive understanding of the role of ESL in pricing, it is necessary to present a brief history of the universe this technique attempts to join. Price tags first evolved from being absent or coded, an approach that allowed retailers to negotiate prices with each client (Spellman, 2009), to being open, which both coped with and promoted self-service. The beginning of the 20th century witnessed the introduction of four generations of devices. The first devices were simple price cards, often handwritten and used mostly to appeal to consumers in shop windows. These devices were soon supplemented with clamping tags, swing tags, and price tag moldings, which became popular in stores between the 1920s and the 1930s (Cochoy et al., 2016a). Fig. 1. Price cards, clamping tags, Swinging tags, and price tag mouldings © The Progressive Grocer, February 1937, p. 103 These technologies were solutions designed to master price fixing operations. It is crucial to note that the term “price fixing” conveys three different meanings. First, price fixing involves -8- price setting: once prices are displayed, they become ultimatums. In other words, they mark a transition from invisible and individually adjustable (i.e. bargaining-based) prices or prices ceaselessly adjusted along market moves to visible “administered” prices that must be either taken or left. Price fixing supports the logics of price formulas, mark-up schemes, and price strategies (Callon, 2013), in which the price is no longer the result of instant power relationships among the innumerable actors of supply and demand, but is instead the outcome of a systematic calculation procedure implemented at the local level. Price fixing must also be understood as price repairing: the growth of consumer markets and the resulting competition among independent grocery stores, chain stores, and supermarkets creates an unstable environment that challenges price formulas, requires their continuous revision, and ultimately leads to continuous price updates. Price tag technologies were invented precisely to overcome the tension between the two first meanings: price fixing as price setting and price fixing as price repairing. They produced third meaning of fixing, this time in the material sense of price clamping: clamping systems were invented by price tag manufacturers to allow prices to be fixed to shelves while remaining moveable; swinging tags were later introduced to combine the permanent display of prices with the easy removal of the goods behind them; and, finally, price tag moldings and astute “clipping” systems were devised to simultaneously allow prices to be slid off and removed and also prevent consumers from displacing them. ESLs are the latest device introduced in this long series of innovations, and as such, they carry not only new promises in terms of increased changeability and readability, but also new uncertainties in terms of their capacity to fit (or not) with the existing “agencement.” In order to understand how ESLs fit into the history of pricing and what changes they convey, we propose a shift from the past to the present and from the general history to a local case by examining how ESLs from one specific provider, the Pricer company, are used in ICA retail outlets in Gothenburg. On its website, Pricer provides six key arguments in favor of ESLs.2 First, ESLs support a complete elimination of discrepancies between shelf prices and checkout prices; second, they reduce the labor time required to change prices; third, they are more reactive to prices; fourth, they ease the implementation of attractive discounts and offers; fifth, they support store management by displaying product information directly from the shelves to the shelf edge (e.g. for out-of-date products); and sixth, they can adapt to different needs and support future updates. A closer look, however, reveals that these arguments are two-sided: on one hand, they are foremost business-to-business arguments, focused on preventing pricing error discrepancies, reducing labor costs and increasing productivity; on the other, they are also consumer-oriented, since reducing errors increases consumer trust, supports consumer choices by improving visibility, and allows stores to offer more attractive discounts through increased flexibility. In fact, we can combine the two rhetorics according to the “B to B to C” pattern we identified elsewhere (Cochoy et al., 2016b). Specifically, by foreshadowing the consumer appeal (B to C) of improved price management (B to B), Pricer’s arguments may intrigue professionals even further, increasing the chances that retailers will buy devices that supposedly increase consumers’ willingness to purchase (B to B to C). Like some of its competitors (e.g. DigiSystem and SES-imagotag), Pricer offers two main label ranges: segment and graphic (see Fig. 2). On the ESL market, both graphic and segment can be based on either LCD display technology or electronic paper technology (e-paper). Pricer offers segment labels that are LCD-based and graphic labels implemented on e-paper. 2 www.pricer.com/en/Benefits-of-Electronic-Shelf-Labels/ (accessed 08.06.17). -9- Therefore, Pricer’s “SmartTAG segment labels” vary from 172 to 184 segments, depending on the size of the label, while its graphic “SmartTAG HD labels” vary from 110 to 184 dots per inch. It should be noted that both resolutions are compared to the minimum standard of 300 dpi usually required by the publishing industry. Among other noticeable differences, the segment labels last less time (5 years) and are cheaper (40 SEK per unit) than the graphic labels (7 years and 120 SEK per unit). Both use interchangeable batteries. We will explore the differences between and implications of these two models in more detail below. Fig. 2. Segment label and graphic label Pricer Despite these arguments in favor of ESL, many ICA stores remain unequipped. Of the 26 visited ICA stores, only five had implemented ESL. Of the three stores under study, the two that first adopted ESLs (ICA Kvantum 1 and ICA Nära 2) chose to begin with ESLs only for the fruit and vegetable shelves. Given their position as early movers, these stores faced greater uncertainty at the time of their implementation and probably found it wiser to test their system in a small part of their store before risking a full-scale investment. By contrast, the third store, ICA Nära 1, which adopted ESL only recently, introduced ESLs throughout the entire store at once. Pricer has a central agreement with Sweden’s ICA group according to which the chain’s stores benefit from pre-negotiated favorable prices. However, every ICA store remains free to purchase ESLs or not and to choose its own ESL provider. So far, Pricer has equipped all ICA stores with ESLs in Sweden except one, which adopted solutions from SES-Imagotag. The overwhelming preference for Pricer ESLs is easy to understand: 1) Pricer is the market leader in Sweden; 2) ICA has a central agreement with Pricer that allows local retailers to get better prices; 3) ICA retailers tend to follow their colleagues’ choices in equipment, especially when their colleagues are satisfied; and 4) ESLs are usually distributed by local companies that have equipped ICA stores with many other devices over the years. The interviewed retailers confirm such logics: -10- “They [Big Dot, Pricer’s products distributor] are selling other products as well, so I have known them for 20 years or something.” (ICA Kvantum 1 owner); “I chose Big Dot because we already had others [devices], for example, [the] Cash Guard machine and a few other systems we also buy from them.” (ICA Nära 2 owner); “They are the largest on the market. I think that it is the system which is the most integrated with ICA systems. And we also… it’s Big Dot which is selling that system, and we have other things from them: we have our Cashguard; we have… different things, so I have a good relationship with the company and I feel comfortable buying things from them.” (ICA Nära 1 owner) When they envision equipping their stores, ICA owners face a lot of uncertainties. In order to reduce these uncertainties, they tend to rely on the experiences of their social networks. Once decisions are made, retailers experience the implementation of their new devices alone. To account for these experiences, there is no better means than the retailers’ testimonies and observations of how ESLs are put into practice. 5. Multiplicity of price representation in digitalized stores We already mentioned Pricer’s major selling arguments, such as the elimination of discrepancies between shelf prices and checkout prices, reductions in the time and costs required to change prices, better reactivity in terms of price variations, an easier implementation of discounts, possibilities for communication between the shelves and the shelf edge, and the possibility of future updates in terms of consumer interaction. In order to compare advertising rhetoric with actual practice, let us now consider the retailers’ discourse. The retailers’ testimonies largely support Pricer’s arguments. Retailers acknowledge that discrepancies between shelf and checkout prices have been almost inexistent since the implementation of ESLs; they recognize that ESLs have drastically decreased operation costs; and they note that displaying discounts is now much faster: Interviewer: “Have you noticed fewer errors between represented prices and checkout prices since you [implemented] ESLs?” — “Yes, they are non-existent today, basically.” (ICA Nära 1 owner); “I have better control over the prices and it is less work [than] having an employee doing it.” (ICA Kvantum 1 owner); “It is very easy, very easy. […] with the weekly prices we only, we only start a file and all the labels are changed. […] I think we are a better store because not so often the prices are not right” (ICA Nära 2 owner) ESLs seem to have clearly improved the chain of price representations (Hagberg and Kjellberg, 2015). Unlike unequipped stores, in which the link between prices and products is maintained only through a spatial arrangement (Grandclément 2008), equipped outlets double this spatial arrangement with a wireless connection between the label and the shelf edge. As we have mentioned, this allows retailers to achieve greater and faster control over their prices. Several authors have shown how the digitization of working places has changed workers’ practices and identities. For instance, research has shown how computer screens support traders’ calculations (Knorr-Cetina and Bruegger, 2000), how videoconferences reshape courtroom interactions (Licoppe and Verdier, 2013), or, closer to our field, how automated checkouts redefine cashiers’ practices and identities in retail stores (Bernard, 2012). Do ESLs replace people’s work? The answer is yes, but only partially. All of the interviewed store owners declared that their staff dedicated much less time to price-changing activities since the implementation of ESLs. One owner even declared that when one of his employees retired, he decided that ESLs could do her job. “In 2013, we had a woman who made all the price changes and [took] the paper labels out in the stores. But then she retired, so we thought we could either hire a new person to do it or buy the Pricer labels. We chose to buy the Pricer labels because it is quicker.” (ICA Kvantum 1 Owner) -11- However, it would be wrong to conclude that ESLs can fully replace human workers. Instead, they transform human workers and their working habits. The first workers to be transformed are the storeowners, who can now manage in-shelf price representations from their offices. Before, price-setting activities were decoupled: the storeowners fixed the prices to be represented and the employees working in the aisles represented the prices materially with paper tags (although our investigations tell us that many owners also regularly work in the store). Specifically, the owners of equipped stores have software dedicated to digital prices management on their office computers. The software is simple to operate and does not require any particular training: our interviewees confirmed Pricer’s claims of user-friendliness. The next people affected by ESLs are the staff members, who are redirected toward other tasks, such as logistics or consumer services. However, staff do not completely abandon price representation activities: first, they use part of the time gained from manual price changing operations for putting large price signs and promotional displays, and second they are needed to ensure the replacement of shelf references and to display the new corresponding prices. To accomplish the latter, employees are equipped with little controllers that allow them to connect electronic tags with new products by scanning both the barcode on the back of the electronic tag (and referring to this unique ESL) and the barcode of the new item. This creates the price correspondence with the checkout. Finally, employees must engage in some work related to ESL maintenance (e.g. when an electronic tag is not connected to any products [displaying “OFF” on the screen] or is broken or discharged. However, these situations tend to be rare and marginal: failing tags are usually few and are handled quickly, and ESL batteries last for five years. Hence, ESL automation does change the nature of retail work; however, ESLs do not directly replace the workforce. Arguably, ESL also does not replace paper. Indeed, paper price representations are still widely present in equipped stores. These observations challenge the idea that “digitization” means “de-paperization”. Indeed, it takes only a quick look inside one of these equipped stores to realize how much the “jungle of paper” (Grandclément, 2008) has survived the electronic price colonization of stores. If shelf labels have become almost completely electronic, paper continues to be a privileged material for displaying discounts. Almost every observed promotion represented with an ESL was supported by a second representation of the same offer with paper. On the contrary, many promotions represented on paper lacked electronic counterparts (e.g. in the case of aisle ends, where the paper continues to carry price representations alone as it did before the advent of ESLs). Paper prices proliferate on banners, aisle ends, in specific shelves where prices are still displayed on paper tags, in service sections, and even in some regular shelves among the ESLs, often to fill the gap of a missing tag. From the observed stores, we can therefore distinguish three major price representation situations: prices represented by electronic tags only, which is the case for most regular prices; prices represented by paper only, which is often the case for aisle ends and other cases that we describe below; and prices represented both by electronic and paper devices, as is the case for many shelf promotions. Pure electronic price representations appear to be the most common case in stores equipped with ESLs, especially for shelf display. However, a closer look shows that these electronic price representations do not take a single form. As mentioned above, Pricer proposes two main ranges of labels. We have observed that segment labels are used in most store departments, except in the fruit and vegetable section, where graphic labels hold a monopoly. -12- Paper price representations are still widespread in equipped stores. If electronic price representations are available in two main forms (segment and graphic) and eight possible sizes, paper labels are definitely more heterogeneous and still ubiquitous. Banners and aisles ends remain equipped almost exclusively with paper. In some specific shelves, such as the bread and cake shelf or seasonal shelves, prices can only be displayed on paper, as written chalk representations, or on slates in manual service counters. Finally, we found than many prices displayed in electronic form are supplemented by paper labels. These paper labels can display either prices that differ from those displayed by the electronic tags for the same products (e.g. for multiple unit price promotions) or the same prices (e.g. for repeating and enhancing a promotion). 6. When ESLs meet paper displays As shown in the previous section, many prices are represented by both electronic and paper devices. How can we account for this cohabitation? Do retailers rely on paper because ESLs are not as efficient as their providers pretend or because they see the duplication of prices as a way of enhancing consumer appeal? In the following three subsections, we outline three forms of interaction between paper and ESLs. 6.1 Product flexibility We saw above that different ESL systems can be used to adjust to different conditions: for packaged goods with large cardboard surfaces for information display, cheaper segment labels focused mostly on price may be sufficient, while, for bulk vegetables, sophisticated and more expensive graphic labels may compensate for an absence of paper by providing more information. According to the store managers and the Big Dot salesman, two main reasons explain this specialization. First, retailers are bound by Swedish law to communicate more information for fruits and vegetables than for other groceries (specifically in relation to the products’ origin and classification). Second, fruits and vegetables are subject to changes in both price and quality. This higher frequency of changes favors the use of more sophisticated devices: “When it comes to ordinary groceries, the main thing you do is to change prices. But when it comes to fruits and vegetables and to meat, you have to… hum, tell the consumer from which country it comes, and if it is good… and what kind of classification [there is] on the vegetables. If it is a first- or a second-class vegetable, which means that if you sell tomato from Holland one day and tomato from Sweden the next day, you have to change from where it comes on the labels.” (Big Dot Salesman). Segment labels, due to their less flexible graphic abilities, cannot display all of this mandatory information. Instead, the names, brands, and barcodes of these products are displayed elsewhere, such as with a little sticker glued horizontally to the bottom of the segment label. The fruit and vegetable references are very variable. Retailers using segment labels for these products, therefore, would need to change the related stickers every day or spend time and money printing classic paper tags. Furthermore, with respect to expiration dates, one storeowner declared that she would rather put stickers on the products themselves in order to make the expirations “clearer for the consumer.” This shows that ESLs are not necessarily more flexible or clearer than paper, since such labels cannot differentiate between two products that are the same but have different expiration dates. In other words, while ESLs may be more flexible for changing prices, paper seems to be more flexible for changing other product characteristics. -13- Finally, graphic labels are believed to look better to consumers. These labels differ significantly from the segment labels, whose monochrome, black and grey, low-definition screens look old-fashioned compared to the colorful and hi-resolution electronic devices people use today (e.g. TVs, smartphones, tablets, laptops, etc.). Graphic labels based on e- paper technologies fill this gap. They look remarkably like paper, to the point that during our very first observations, we overlooked their presence in stores and confused them with traditional paper labels. Here lies a paradox: with graphic labels, the difference between traditional paper price representations and new electronic ones is blurred by electronic devices designed not to appear electronic. 6.2 Dramatizing promotions From a marketing perspective, merchandisers often dramatize the presentation of product display. Whereas ESLs present several distinctive back-office advantages, retailers agree that electronic labels are not very attractive in terms of the consumer front end. As has already been mentioned, segment label resolutions can appear old-fashioned; in other words, “It doesn’t look really 2016” (ICA Kvantum 2 owner). From this perspective, one could say that if ESLs spread despite their lack of consumer appeal, it is because the detachment of prices from goods has become standard: price stickers have disappeared, and many consumers pick up goods without closely looking at prices as part of their routine purchasing behavior. In this respect, prices become less important, rendering sad and grey ESLs more acceptable. ESLs’ lack of attractiveness/readability may even encourage such behavior by indirectly focusing consumers more on colorful packaging, ensuring that their interests match those of retailers. Retailers prefer graphic labels, but they are still limited by a lack of color options (e.g., red is becoming increasingly available, but after a while it can fade and turn pink). Why do these screens look so old-fashioned? This seems to be mainly a question of money and power: our smartphones’ color screens are much more costly and energy-consuming than what retailers can afford for their labels. In this context, a label’s battery capacity is a very important issue, since changing the batteries of several thousand tags every week would be far too costly in terms of time and money. Such a situation would involve the same tedious operations required to maintain paper tags, but at a much higher cost. By contrast, paper labels require no batteries once they are printed. Furthermore, they offer proven efficiency, larger sizes, and more attractive colors (or more prisigare, to quote a neologism used by ICA to refer to a better form of price). Another key issue is that of visibility, which is relevant for several reasons. Scholars have shown that visibility is important for understanding the economy because, in modern markets saturated with information, products compete to attract consumers’ attention. Prices serve as “captation” devices designed to increase visibility in stores saturated by their representations (Grandclément, 2008; Hagberg and Kjellberg, 2015). Most ESL provider websites stress the clarity of ESLs as a marketing argument, using such words as visible, clear, high resolution, flashing, readability, noticeable, and contrast.3 Unlike these websites, we do not seek to show how ESLs increase the visibility of prices and product information; we only wish to explore how such visibility is achieved through the combination of traditional paper price representations and electronic ones. 3 Lexical choices noted by observing the websites of Altierre, DigiSystem, DisplayData, Pricer and Store- Electronic-System. -14- Retailers can use paper and electronic tags in combination to multiply the hooking points of prices and promotions (e.g. by emphasizing a price discount; Fig. 3), to promote multiple-unit price offers (Fig. 4), or to stress the price advantages offered to loyalty card holders. Even when electronic prices are not supported by paper equivalents, paper is often used to complement electronic tags in order to signal a promotion (e.g. “Eko prices”) or display more information about products. . Fig. 3. Paper and electronic labels displaying twice the same price for the same product ICA Nära 1, December 2015 Fig. 4. Different ESL and paper prices for the same item ICA Nära 1, December 2015 Double material displays are often characterized by a perpendicular opposition between paper and electronic representations. These “lateral flags” (Cochoy, 2007a) complement price tags and ensure that consumers can always see the product price, whether they face the price tags directly or see the flags as they walk along the shelf. These different prices contribute to the -15- process of consumer “captation,” or non-forceful attraction (Cochoy, 2007b). Indeed, paper flags placed perpendicular to the consumer path attempt to divert consumers from their trajectory, increasing the ability of packaging and electronic tags to portray products and their characteristics. 6.3 Multi-channel marketing Front-office features are complemented by back-office advantages. In particular, it is interesting to consider how electronic tags create new logistical opportunities that improve ICA’s multi-channel marketing system. Despite their independence, ICA retailers change several labels every week in order to stay consistent with the “weekly prices” mandated by the national ICA campaigns. These “weekly prices” are displayed via TV ads, on the Internet through the ICA website and Facebook page, in the customer magazine Buffé (Ots, 2010), and in paper magazines provided at store entrances (in this case, generally in cases of “weekly prices” related to specific store formats). Thus, on Monday morning, retailers have to change all the “weekly prices” from the previous week back to ordinary prices and to display the promotions of the new week, and they need to perform this change before the store opens. Thanks to electronic labels, equipped retailers are released from the time-consuming work of changing “weekly prices” on Monday mornings, allowing them to instead focus on filling shelves and putting up large paper signs. Therefore, we can argue that ESLs improve the ICA multi-channel marketing system in at least three ways. First, they increase its temporal consistency, since labels can be updated as soon as the store opens. Second, on Monday morning, retailers have more time to focus on the installation of different paper devices promoting weekly discounts. Ironically, the use of ESLs, instead of easing the substitution between electronic and paper devices, actually supports the implementation of the latter. “In the history, we had to put up new tags every week for the discount prices; now we don’t need that— we can just put bigger signs” (ICA Nära 2 owner) Third, by winking, the electronic tags cannot only divert the trajectories of strolling consumers, but also facilitate the work of those coming for specific promotions advertised in external media. Hence, by taking place in an already structured chain of prices, electronic tags not only collaborate with paper labels in store, but also improve ICA’s multi-channel marketing system on a broader scale. In the case of promotions displayed on paper in electronically equipped stores, global displays support a semiotic opposition between the electronic tags and the paper promotions. The grey, low-contrast electronic tags contrast sharply with the diverse array of bright colors of paper tags, and this contrast increases the visibility of the latter. Furthermore, the divisions among self-service shelves (priced with segment labels), service departments (priced with paper or chalk signs) and vegetable sections (priced with graphic labels) improves the clarity of store organization by adding visual differences that reinforce at the shelf level the store’s thematic organization, which is often displayed on large ceiling signs (Cochoy, 2007a). However, it seems that these organizational effects of various ESLs and price arrangements are not intentional. As mentioned above, the division between vegetables and other shelves is mainly the result of logistic constraints and legal requirements. However, this fortuitous (rather than strategic) opposition shows how much technical objects can spark unexpected processes and take on roles beyond those intended (Akrich et al., 1988). Today, it seems that paper and electronic prices achieve different purposes: ESLs are great logistical improvements because they increase retailers’ control over prices. However, for several reasons related to power and cost limitations, they are still less attractive than paper -16- price representations when it comes to the emotional aspect of prices. In other words, whereas ESLs primarily serve professional back-office needs, paper price signs address consumers. “The electronic label is not a tool to sell more. The electronic label is a tool to show you the right price at the right time. […] So we are not… what do you say, competitive… We are not competing for making offers. We are competing for showing the right price when it is the right price.” (Big Dot salesman) How could we explain this situation? One reason is that electronic tags are not necessarily more flexible than paper labels. Though ESLs have greater flexibility for representing prices in numeral terms, they sorely lack flexibility for representing prices as pictorial signs requiring various supports (cf. Grandclément, 2008): they have no colors, they are constrained to a standardized and rigid design4, there are only few formats dedicated to specific products, they cannot be put everywhere, etc. Meanwhile, paper is easy to design, easy to print, and easy to handle. The problem retailers face is that, in stores, a price is preferably both arithmetically and pictorially flexible. Achieving this dual flexibility is not possible with a single device, since paper and electronic solutions are limited by irreducible material properties. Despite their different roles, by working together, paper and electronic solutions can turn their respective weaknesses into strengths to improve multi-channel logistics, increase the visibility of promotions, and enhance consumer attraction. 7. Putting Gothenburg ESLs in perspective Our results convey several insights, but to be fully meaningful, these insights should be put into perspective. As we argued, Pricer’s ESLs in Gothenburg are among the most advanced solutions at the European level, since Pricer is the leader in the European market. However, our case study is not sufficient to gain a global perspective on the potential and limits of ESL technology. To put our results in perspective, we propose contrasting our case with one from a distant time and place: the American case we started from. This comparison is particularly useful because it reveals puzzling and counter-intuitive clues. Let us first go back to the American history of price tags. Hindsight supports some ambivalent observations. On one hand, it shows that ESLs appeared very early in the US, probably earlier than they did in Europe, since the first experience reported by The Progressive Grocer goes back as far as 1985 (The Progressive Grocer, November 1985, p. 54). However, on the other hand—and surprisingly—a systematic reading of the same magazine shows that ESLs have not made much progress on the American market: though ESLs occasionally gain popularity, their coverage is by no means comparable with that of more popular technologies, such as loyalty systems, POS software, and so on. This is confirmed by a second shift in perspective, this time in terms of geography. Thanks to a quick ethnography conducted in New York retail outlets between late March and early April of 2016, we discovered a surprising reality. Though we found no store equipped with ESLs in New York City, we did observe a widespread use of highly sophisticated paper shelf stickers. These stickers were impressive in terms of number, size, and pricing strategies, and their locations on the shelves allowed them to mask or extend existing price tags by proposing batch sales, highlighting advertised products, offering thousands of points to card loyalty members, introducing comparisons between product leaders and the store’s private brand, 4 Pricer proposes a standardized design for ICA. Though most ICA retailers follow this design, they remain free to choose other ones. -17- suggesting “smart buys,” urging consumers to seize bargains during weekly sales or take their “last chance,” and so on (see Fig. 5). Fig. 5. Beyond ESL: Price stickers in New York New York, April 2016 In a sense, many would conclude from the lack of ESLs in American retail settings that US retailers are “late” in terms of price display techniques, at least compared to Sweden. For once, Europe seems to take the lead over the other side of the Atlantic! Consider Pricer: the Swedish ESL manufacturer is successful everywhere except on the American market, which appears to lag behind in terms of ESL adoption. However, the paper stickers dear to the US retailers remind us that “late” is a dangerous word. Although descriptive at face value, this adjective, when considered seriously, is, in fact, fully normative and prescriptive: it implicitly -18- ordinates what is observed by deciding what represents the future or not, and it supposes that some elements are inherently moving somewhere and that what does not move as fast or in the same direction loses a sense of history. Digitization can be seen as one manifestation of this rapid and one-directional moving process. The idea that digitization points towards a necessary evolution is shared by the providers of digital devices, whose duty is precisely to position their products as anticipating the next state of affairs. This view is often shared by their clients, who are seduced by the promises of new devices and afraid to stay behind in what they perceive as being irresistible transformations. Last but not least, the modernist view is even easily adopted by social scientists, who, by considering just one element of their field, may overlook the larger picture. Now, the sophistication of price stickers tells another sorry: it reminds that paper is about “heating” consumer places (Grandclément, 2008). If graphic labels look like paper, they are rigid. By contrast, fragile stickers remind the swinging tags of the past (see the first image of Fig. 5, in which some stickers are clearly swinging backwards). Paradoxically, even if electronic devices are the fastest conveyors of price information, soft, ephemeral papers are better able to give the impression of hot, short-term bargains that one should seize by impulse. Moreover, paper stickers function as a conspicuous testimony of back-office work; they show the service behind self-service, and thus serve as a means of paying attention to consumers. Last but not least, these paper price tags are part of a more general system in which the legal obligations of item pricing have long impaired the development of electronic solutions (Cochoy et al., 2016a). If, as social scientists, we do not need to decide which device is the most appropriate or do not have the means to fully assess the efficiencies of different technologies, we can at least stress their respective features and outline that the success of one over the other is far from guaranteed. The uncertainty about which of the two devices matters most is all the more significant because, as we have shown, both approaches prove more effective when articulated rather than opposed: while ESLs can surely help grocers in their back-office operations, paper displays still have words to say—or, rather, numbers to tell and colors to show—when it comes to attracting consumers’ attention. The fewer ESLs there are, the more sophisticated the papers will be. Innovations should not be made in isolation: what matters is neither ESLs nor paper tags, but the complex and wider price agencement to which they both belong. 8. Conclusion: a matter of matter ESL is a major retail innovation, but in the stores where this innovation has been introduced, paper price representations have not disappeared. This study has identified three approaches to price representations: prices represented by paper only, prices represented by ESL only, and prices represented by both paper and ESLs. By considering these approaches within the the theoretical frameworks with which we began, we can perceive that these different alternatives convey at least three lessons. The first lesson, as we just saw, is that price technologies should be considered not in isolation, but as part of a complex web of practices and market agencements. What matters is not the property of the technology itself, but the performance of the overall system to which it belongs. When considered alone, ESL technology appears to be a rational novelty with superior features that should drive old-fashioned paper competitors out of the market. However, when considered within the wider price agencement, ESL appears to be one element among many; it offers several benefits, but also suffers weaknesses. Thus, ESLs must “co-operate” with other elements, each with advantages and drawbacks, to improve overall -19- performance. What matters in this case is not the features of any given component, but the quality of their combination. A second lesson is that prices should not be seen as the results of abstract encounters of mysterious global entities like supply and demand, but as the outcomes of complex market infrastructures. For instance, by easing price changes and giving managers full control over such changes, ESLs help to introduce an asymmetrical pricing strategy that produces fast changes when supply prices go up and slow changes when they go down. The third lesson is that ESLs support a new conception of prices based on the idea of greater price flexibility and transparency. According to Austin (1970), performative utterances are “acts of language” that have the ability to simultaneously describe and enact a state of the world. One such statement is “I promise that my pricing system will improve your business.” Callon (1998) showed that market theories are performative and that they are performed through ad hoc market devices. However, the same author stressed that the performances of economic conceptions are often subject to “misfires” (Callon, 2010). This means that performative statements often fail due to their intricacies and sociotechnical agencements, which hamper, bias, and distort their messages. As we have seen, ESLs do not escape this rule: for instance, their claims are frequently challenged by both their inner limitations (e.g. small size, sad color…) and the properties of the other elements with which they work (e.g. more colorful and talkative paper, traditional products, and so on.). In sum, we now understand that the performance of ESLs depends not only on the promises of the devices themselves, but also on their ability to properly fit with all of the actors, tools, discourses, objectives, and expectancies that are part of the overall retail agencement. Acknowledgements We warmly thank The Progressive Grocer for granting us the permission to reproduce the image displayed above as figure 1. We also thank the two anonymous reviewers for their constructive comments and suggestions. This work was supported by the Swedish Research Council within the framework of the Digcon Project [Grant number 2012-5736]. References Ahmetoglu, G., Furnham, A., Fagan, P., 2014. Pricing practices: A critical review of their effects on consumer perceptions and behavior. Journal of Retailing and Consumer Services. 21 (5), 696-707. Akrich, M., Callon, M., Latour, B., 1988. À quoi tient le succès des innovations ? 1 : L’art de l’intéressement; 2 : Le choix des porte-parole. Gérer et Comprendre. 11, 4-17 & 12, 14-29. Austin, J.L., 1970. Quand dire, c'est faire. Editions du Seuil, Paris. Baiocchi, G., Graizbord, D., Rodríguez-Muñiz, M., 2013. Actor-Network Theory and the ethnographic imagination: An exercise in translation. Qualitative Sociology. 36 (4), 323–341. Bernard, S., 2012. Travail et Automatisation des Services. La Fin des Caissières ? Octarès, Toulouse. Bijker, W.E., Law, J. 1992. Shaping Technology/Building Society: Studies in Sociotechnical Change. MIT Press, Cambridge, MA. Biswas, A., Dutta, S., Pullig, C., 2006. Low price guarantees as signals of lowest price: The moderating role of perceived price dispersion. Journal of Retailing. 82 (3), 245–257. Borden, N.H., 1964. The concept of the marketing mix. Journal of Advertising Research. 4, 2–7. Bowker, G.C., Baker, K., Millerand, F., Ribes, D., 2010. Toward information infrastructure studies: Ways of knowing in a networked environment, in: Hunsiger, J., Klastrup, L., Allen, M. (Eds.), International Handbook of Internet Research. Springer, Dordrecht, pp. 97-117. -20- Callon, M., 1986. Some elements for a sociology of translation: Domestication of the scallops and the fishermen of St-Brieuc bay, in: Law, J. (Eds.), Power, Action and Belief: a New Sociology of Knowledge? Sociological Review Monograph: Routledge and Kegan Paul, London, pp. 196-223. Callon, M., 1998. The Laws of the Markets. Blackwell, Oxford. Callon, M., 2010. Performativity, misfires and politics. Journal of Cultural Economy. 3 (2), 163-169. Callon, M., 2013. Qu’est-ce qu’un agencement marchand ? in: Callon, M., Akrich, M., Dubuisson-Quellier, S., Grandclément, C., Hennion, A., Latour, B., Mallard, A., Méadel, C., Muniesa, F., Rabeharisoa, V. (Eds.), Sociologie des Agencements Marchands. Presses des Mines, Paris, pp. 325-440. Callon, M., 2016. Revisiting marketization: From interface-markets to market-agencements. Consumption Markets & Culture, 19 (1), 17-37. Callon, M., Muniesa, F., 2005. Economic markets as calculative collective devices. Organization Studies. 26 (8), 1229-1250. Callon, M., Millo, Y., Muniesa, F., 2007. Market Devices. Blackwell, Oxford. Cochoy, F., 2007a. A sociology of market-things: On tending the garden of choices in mass retailing, in: Callon, M., Muniesa, F., Millo, Y. (Eds.), Market Devices. Blackwell, Oxford, pp. 109-129. Cochoy, F., 2007b. A brief theory of the “captation” of publics: Understanding the market with little red riding hood. Theory, Culture & Society. 24 (7-8), 203–23. Cochoy, F., 2016. On the Origins of Self-Service. Routledge, Abingdon, Oxon. Cochoy, F., Hagberg, J., Kjellberg, H., 2016a. Displaying, sticking and digitizing prices. Paper presented at the 4th Interdisciplinary Market Studies Workshop, St Andrews, UK. Cochoy, F., Smolinski, J., Vayre, J.S., 2016b. From marketing to “market-things” and “market-ITing”: Accounting for technicized and digitalized consumption, in: Czarniawska, B. (Eds.), A Research Agenda for Management and Organization Studies. Edward Elgar, Cheltenham, UK, pp. 26-37. Falk, T., Kunz, W.H., Schepers, J.J.L., Mrozek, A.J., 2016. How mobile payment influences the overall store price image. Journal of Business Research. 69 (7), 2417–2423. Grandclément, C., 2008. Vendre sans Vendeurs: Sociologie Des Dispositifs D’achalandage En Supermarché. Thèse de Doctorat, École Nationale Supérieure des Mines, Paris. Grewal, D., Janakiraman, R., Kalyanam, K., Kannan, P.K., Ratchford, B., Song, R., Tolerico, S., 2010. Strategic online and offline retail pricing: A review and research agenda. Journal of Interactive Marketing. 24 (2), 138–154. Grewal, D., Ailawadi, K.L., Gauri, D., Hall, K., Kopalle, P., Robertson, J.R., 2011. Innovations in retail pricing and promotions. Journal of Retailing. 87 (1), S43–S52. Hagberg, J., 2016. Agencing practices: A historical exploration of shopping bags. Consumption Markets & Culture. 19 (1), 111-132. Hagberg, J., Kjellberg, H., 2015. How much is it? Price representation practices in retail markets. Marketing Theory. 15 (2), 179-199. Hagberg, J., Normark, D., 2015. From basket to shopping bag: Retailers’ role in the transformation of consumer mobility in Sweden, 1941-1970. Journal of Historical Research in Marketing. 7 (4), 452-475. Hagberg, J., Sundstrom, M., Egels-Zandén, N., 2016. The digitalization of retailing: An exploratory framework. International Journal of Retail & Distribution Management. 44 (7), 694-712. Hall, J.M., Kopalle, P., Krishna, A., 2010. Retailer dynamic pricing and ordering decisions: Category management versus brand-by-brand approaches. Journal of Retailing. 86 (2), 172–183. Hyysalo, S., Jensen, T.E., Oudshoorn, N., 2016. The New Production of Users: Changing Innovation Collectives and Involvement Strategies. Routledge, New York. Knorr-Cetina, K., Bruegger, U., 2000. The market as an object of attachment: Exploring postsocial relations in financial markets. Canadian Journal of Sociology. 25 (2), 141-168. Kopalle, P., Biswas, D., Chintagunta, P.K., Fan, J., Pauwels, K., Ratchford, B.T., Sills, J.A., 2009. Retailer pricing and competitive effects. Journal of Retailing. 85 (1), 56–70. Latour, B., 1992. Aramis, ou l’Amour des Techniques. La Découverte, Paris. Latour, B., 2005. Reassembling the Social. Harvard University Press, Cambridge, MA. Licoppe, C., Verdier, M., 2013. Interpreting, video communication and the sequential reshaping of institutional talk in the bilingual and distributed courtroom. The international Journal of Speech, Language and the Law. 20 (2), 247-276. -21- Lombart, C., Louis, D., Labbé-Pinlon, B., 2016. Price image consequences. Journal of Retailing and Consumer Services. 28 (1), 107-116. Marcus, G.E., 1995. Ethnography in/of the world system: The emergence of multi-sited ethnography. Annual Review of Anthropology. 24, 95-117. Muniesa, F., Millo, Y., Callon, M., 2007. An introduction to market devices. The Sociological Review. 55 (2), 1–12. Ots, M., 2010. Understanding Value Formation. A study of Marketing Communications Practices at the Food Retailer ICA. PhD Dissertation, Jönköping International Business School, Jönköping. Oudshoorn, N., Pinch, T., 2003. How Users Matter: The Co-Construction of Users and Technologies. MIT Press, Cambridge, MA. Shove, E., 2010. Beyond the abc: Climate change policy and theories of social change. Environment and Planning A. 42 (6), 1273-1285. Shove, E., Pantzar, M., Watson, M., 2012. Dynamics of Social Practice. Sage, London. Spellman, S., 2009. Cornering the Market: Independent Grocers and Innovation in American Small Business, 1860-1940. PhD Dissertation, Carnegie Mellon University, Pittsburg, PA. Von Hippel, E., 2005. Democratizing Innovation. MIT Press, Cambridge, MA. Web References ESL provider websites: http://www.altierre.com (accessed 08.06.17) https://www.digisystem.com (accessed 08.06.17) https://www.displaydata.com (accessed 08.06.17) http://www.ses-imagotag.com (accessed 08.06.17) http://www.pricer.com (accessed 08.06.17)