As with other information technology (IT)-enabled service sectors, the evolution of the retail sector occurred in stages. As retailers grew in size, they first used digital tools to capture information and routinize processes, such as checkout, ordering or stocking. Next, more sophisticated demand management tools began analyzing the data streams from point-of-sale (POS) databases, opening the opportunities to re-organize shop floors, supply chains, and find complementarities (McKinsey 2001).Footnote 8
We can identify several major stages of the digital revolution in retail trade. While the precise timing of each varies cross-nationally, the first stages began in retailing during the 1970s and expanded in the 1980s while shifts in value capture began in the 1990s and continue today.
The earliest stages involved the use of technology to increase economic and information power. Retailers first used barcode technology, scanning system, and basic purchasing software to codify information and routinize activities. Next, they began using more complicated merchandise planning and demand management software to begin organizing and analyzing the stream of sales data routinization produced. Along with complementary store changes (such as decreased warehouse space and increased sales floor space) these technologies allowed larger stores, larger store networks, increased product diversity and knowledge, all while reducing non-sales inventories. Increased scale led to clashes with small shopkeepers and national politicians around store size, location, purchasing power, price rules, and anti-trust and competition issues.
The marriage of technology and scale led both new business possibilities and contests over information use. Once retailers had implemented basic digital technologies and built large-scale retailer networks, they sought to use both their informational and economic power to capture new value and offer new services. IT plus improved data allowed increased supplier integration, network rationalization, customer niche targeting, efficient customer response (ECR) technologies, loyalty programs, reduced risks with private labels, worker up-skilling, and more. Opportunities for business model extensions included offering new services in store (financial, basic medical, food services, advertising, etc.), controlling and streamlining distribution and wholesaling, or creating and branding products. Competition between retailers and with suppliers and wholesalers has created political clashes around data use (privacy), supplier-relations (including terms of payment), and even what services retailers can offer.
The 1960s was not the beginning of large retailers, as enormous retail firms have existed for some time. However, digital technology has changed the retail market from one where large firms were the exception to one where they are the rule. It has done so by reducing the risks associated with scale, increasing the advantages of that scale, and providing tools for scale management.
For example, in 1945, sales at Sears and Roebuck, the largest US retailer, passed $1 billion (around 0.45% of US GDP or $11.90 billion in 2009 equivalent dollars).Footnote 9 Sears struggled with its size. In 1946, General Wood, then president of the company noted its problems, stating, “I have been in an unusual position to observe the problems that ‘bigness’ brings to a business” (Emmet and Jeuck 1950, p. 365). The solution was massive decentralization, breaking the firm into five almost wholly autonomous territories, and within each stores still had considerable flexibility. In 1964, a Forbes article decreased Sears as “number one in the U.S., and also number 2, 3, 4, and 5” (Katz 1987, p. 15). Even with its successful decentralization, Sears was the anomaly, able to succeed at its size only because it set the market, telling consumers what to buy rather than simply providing it, helping reduce demand management risks.
Fast forward to 2009. That year, Walmart totaled $406 Billion in sales, the equivalent of 2.84% of GDP ,Footnote 10 six other retail firms (Home Depot, CVS, Kroger, Costco, Target, and Walgreen) had revenue equivalent or larger to Sears’ 1945 level of GDP, another five (Sears, Lowe’s, SuperValu, Best Buy, and Safeway) had sales above $40 billion, and eleven more had revenues at or above $11.90 billion, the 2009 dollar equivalent of Sears’ 1945 level.Footnote 11 In dollar equivalents, all 23 of these firms would have been America’s largest retailer in 1945. How was this explosion of enormous retail empires possible? The answer begins with the barcode.
Digital information has fundamentally changed retail firm size by allowing retailers better information about sales, better response times, better communication with store networks, and reduced risk through better demand management. The beginning of the scale revolution in retail trade was the digital routinization of basic activities. Prior to the full onset of the digital era, earlier technological advances, such as electronic registers had moved the sector partially in this direction. With the implementation of barcode technology (see Nelson 1997), scanners, and the digital storage of sales and stock information, however, in-store operations and automation jumped forward. These basic digital technologies allowed the automated gathering and storage of information provide the basis for each of the other stages described below. The fundamental feature of the digital revolution is the ability to create, store, transfer, and manipulate digital information, and barcode technology is the point at which digital information is created. Basic digital product and sales information then cascades through the rest of these stages.Footnote 12
The ability to store information about product price and stock levels through barcodes revolutionizes store operations. In addition to accelerating the movement toward self-service in retail, stores are able to reduce inventory checks (a timely and costly procedure), speed-up checkout times, and generally create the types of innovations toward larger-volume formats, reaping the benefits of greater scale while keeping organizational challenges in check.Footnote 13 In addition, IT has also provided a valuable tool for headquarters to effectively manage large chains of stores. IT allows a central location to communicate with stores, gather the incredible stream of information from various locations for analysis, and coordinate larger distribution networks more seamlessly. Data on electronic data interchange systems (EDI), which are used to automatically place orders with suppliers, shows the increasing routinization of retail. Hwang and Weil (1998) find that the use of EDI by retail firms increased from 33 to 83 percent in the period from 1988 to 1992 alone.
Once product information is digitized, retailers can begin analysis of sales information, allowing superior management of both inventories and price decisions. Retailing occupies a critical economic location between manufactures and consumers. For retailers, therefore, understanding what products consumers demand and at what price is a fundamental component of providing retail services. While barcode technology allowed retailers enormous labor savings, equally large savings have accrued from the analysis of the point-of-sale (POS) data that barcodes generate. Since the mid-1990s, nearly every major global retailer has kept data on every purchase made across their store networks, often using loyalty cards to tie these purchases to particular customers.Footnote 14 The analysis of this data allows firms to better understand and manage consumer demand in a number of ways including: better matching of inventory to customer demand, even below the product level to features such as color, better understanding of the relationship between prices and sales, more efficient use of shelf space and more efficient store layout through recognizing sale complementarities, reduced inventory and fewer out-of-stock situations, and the potential to evaluate and optimize advertising, even at a personal level.
Again, this is not simply about cost cutting, but also about improving sales through better store layouts, keeping goods on the shelves, and generally matching demand. Dobson et al. (1998) find that in the United Kingdom between 1980–94 retail sales (in real terms) per outlet increased by 53% and per employee by 23%.
Technology also allows reduced overheads and smaller stock warehouses, which can be seen in data on assets and liabilities in retail trade over time. Looking at French retailers, Dawson (2005) finds that assets and liabilities were down to 47.4% and 53.6% of sales, respectively, in 2001. This is compared to 63.1% for assets and 59.4% for sales in 1984. Gaur et al. (2005) support this finding from a new direction, demonstrating that greater capital intensity leads to increased inventory turnover (and lower margins). In other words, investing in information technology has helped retailers to further increase turnover and lower inventories relative to sales, increasing returns on capital. As a percentage of sales, retailers can hold on to fewer goods and incur fewer liabilities as a percentage of assets. Taken together, therefore, effective demand management allows productivity gains on capital, space, and labor. The ability to keep items in stock allows storage space to be reduced or even eliminated, increasing productivity in terms of sales per square foot. Buying products from suppliers as they are needed cuts down on overstocked inventory, increasing capital productivity. Finally, retailers can now place goods in complementary locations, carrying more of what consumers want placed by other goods they buy them with, which increases sales of both.
To recapitulate, digitally captured sales information and analysis allowed retailers to routinely achieve scale and use scale advantages to handily out compete small independent retailers without similar capacity. Holmes (1999) report for the Federal Reserve Bank of Minneapolis summarizes many of these findings. It models the adoption of new information technologies in the retail industry, arguing that barcodes and increased delivery offered a variety of new business strategies built around increasingly large retail networks.Footnote 15
Increasing business opportunities have been multiplied by similar IT-transformations of industries that complement retail, such as transportation. On-board computers (OBCs) and electronic tracking has increased capacity utilization and tied transportation more tightly into firm networks. This follows the expectation of Bresnahan and Greenstein (1997), who argue that productivity growth in services requires co-invention across firm networks.Footnote 16
Once firms had scale and a steady information stream about what customers were buying, when they were buying it, and under what store and price conditions, they faced an array of new business challenges and opportunities. How should they integrate technology into existing patterns of labor implementation—should they use information to automate or enable workers? Should they share sales data with their suppliers? How should they gather information from and connect with customers to better individualize service provision? Finally, what additional value opportunities were available? Which should be seized with partners and which should be captured alone?
The answers to these questions are neither simple, nor inherent in the technological capabilities of digital tools. Looking ahead to the outcome, we find that firms have solved them in a multitude of ways. Why? The answer lies in political deals about the distribution of wealth in the economy, as political negotiations pushed firms toward nationally specific calculations about the market, firm, and relational contracting choice for firms poised by Williamson (1985).
Interestingly, many of the divergences in digital strategy appeared before digital technology was central to core retail processes (for a much longer discussion of the political competitions, their national character, and the outcomes sketched below, see Watson 2011). The majority of powerful retailers in the global economy are actually relatively new firms. Unlike many industries, essentially every major global retailer is a recent company, with few leading retailers existing with any size prior to the 1950s.Footnote 17 By the 1960s, however, a new class of retailers had emerged across the affluent economies. These retailers were spurred common social and economic changes in the post-war period and process innovations including self-service, ample automobile parking, larger stores, and more product diversity. In all countries scale retailers emerged in a similar fashion, promoting self-service and the stack it high and sell it cheap philosophy.
The rapid growth in scale retailers, however, soon created clashes with independent shopkeepers, long a fixture of both the economy and the ballot box. Independent shopkeepers wanted protection from what they saw as either unfair or socially disruptive aspects of scale retail development including predatory pricing, out-of-town development, large stores, and their scale more generally. A political conflict was brewing, and both sides began to mobilize coalitions to lobby for their preferred regulatory outcomes. How retailers built these coalitions was quite different by location, motivated by different political challenges from opposition groups of varying power and institutional resources.
Retailers are among the most connected actors in the economy, interacting with a variety of economic, political, and social groups. Politically, they must manage relationships ranging from international trade organizations and national regulators down to municipal governments. Within national economies, they connect with consumers, suppliers and producers, manufacturing firms, and wholesalers.
The development of the three national models of retail capitalism described above is best explained by how retailers built these political partnerships in the 1960s and 1970s. The form and function of each national coalition proved a critical determinant in subsequent nation-specific business strategies; building and maintaining successful coalitions not only forged economic partnerships, it also influenced ongoing regulatory strategies for national retail sectors.
Retailers facing the most fragmented and decentralized national environments (as in the United States) faced the least organized opposition and consequently needed the fewest political partners. They faced the most liberal regulatory environments and formed lean retailing business strategies built on low levels of cooperation with workers, suppliers, and governments. In the most highly consensual and corporatist countries, by contrast, powerful external and internal opposition groups led retailers to build broad political coalitions that germinated relational contracting business models. These business strategies forced retailers to integrate labor and manufacturing partners into business strategy or face political, social, and economic consequences. These coalitions agreed to higher levels of buffering regulation in return for political and economic alliances. Finally, vertical integration retailers emerged in environments with powerful but fragmented opposition groups by mobilizing weak, short-term coalitions where advantageous for development, such as local political partnerships, minor worker concessions, and selective cooperation with industry.Footnote 18
Each coalition re-shaped the economic calculations of retail firms, pushing them toward different strategies of labor relations, supplier connections, product strategies, value added-services and ultimately the digital strategies that now underpin each of these dimensions of business competition. Although it is not in the scope of this article to fully detail each model, the data below provides quick evidence that they are real and sustained.
Comprehensive data on technology strategies are difficult to assess, but data in Table 3 from Kurt Salmon and Associates on Efficient Customer Response, a good proxy for the digital interactions between retailers and suppliers and customers, shows a large difference in the lean retailing US versus European firms.
A similar divergence in product strategy development has emerged, driven by the vertical countries. Data in Table 4 show that not only do the vertical integration countries lead in private labels as a percentage of sales, but there has been little convergence. Sales of private labels in the UK and France were 15% higher than in the US in 1995 and 18% higher in 2005. In addition, these numbers hide qualitative differences in how private label goods are used, primarily simply as low-cost alternatives in the United States versus high value added products at a variety of price and quality points.Footnote 19 Note that private labels are only one dimension of the value capturing strategy of vertical integrators, which also includes additional services in-house and value added products such as fresh prepared meals in food retailing.