3TOWARD THE DATA-POWERED AISLE

During the second week of December 2011, Amazon encouraged shoppers to install the company’s price-check app on their smartphones and then scan the UPC bars of items they were considering buying at local physical retailers. As an enticement, the company offered a 5 percent discount (up to $5) if the customer purchased the scanned item from Amazon (the deal applied only to certain products). Not only could the shopper benefit by paying less, but the app also provided the e-tailer with the location and the price of the product—critical information pertaining to its physical-store competitors throughout the country during the most popular shopping time of the year. With this information Amazon could then adjust its prices by geographic area to achieve optimal profit margins.1

To many merchants, the mix of barcode scanning, location checking, and a discount—on top of no sales tax—reinforced Amazon’s image as a destructive pirate. Of particular concern was Amazon’s encouragement of “showrooming,” which is the practice of inspecting merchandise at a physical store but then making the actual purchase online. The publishing industry has little doubt that this tactic caused the downfall of many U.S. bookstores. The author Scott Turow, president of the Authors Guild, decried the “bare knuckles” approach to retailing that Amazon’s broader price-comparison project revealed.2 Other critics focused on the possible consequences for small businesses. The Sun newspaper in Lowell, Massachusetts, editorialized that “the online behemoth is trying to put small business owners on the street with a Christmas sales promotion that is—quite frankly—unfair and un-American.”³ The New York Times drew a parallel between Amazon’s activity and “when Walmarts open in small towns.”⁴ Similarly, a Canadian bookseller rejected the admiring portrayal of Amazon CEO Jeff Bezos, who transformed the way many purchase books, “as Steve Jobs of the book world. For me, he’s the (Walmart founder) Sam Walton of the book world.”⁵

Large retail chains, including those with online components, were also angry that their major internet competitor was encouraging huge numbers of people to exploit the physical stores they frequented. “Amazon has been at the implicit war with local brick-and-mortar stores” since its inception, wrote New York Times columnist Robb Mandelbaum. “Last week, the implicit seemingly became explicit.”⁶ Target signaled its irritation in a letter to suppliers that promised “what we aren’t willing to do is let online-only retailers use our brick-and-mortar stores as a showroom for their products and undercut our prices.”⁷ The customer “has so many more available places to get products, and mobile technology and the Web have completely exposed pricing,” said Al Sambar, a retail strategist at the consulting firm Kurt Salmon. “So the two best competitive levers retailers have are under attack.”⁸ Five months later Target stopped selling Amazon’s Kindle e-readers.9 The company certainly knew that Amazon’s app was not the only smartphone price-checking software, but officials were especially angry that a giant online merchant was gleefully poaching customers systematically from inside a competitor’s physical store.

The year 2011 signified the beginning of a great transition in America’s retailing institution, as a broad swath of retailing executives expressed consternation over Amazon’s activities and their inability to combat them. Retailers finally opened their eyes to the mountains of data that now could be accessed because of the Universal Product Code, the internet, and other new technologies, and they started to rethink their entire approach to customers, merchandise, pricing, and the selling process. As a result, we are entering a new retailing era with swiftly growing impulses toward customer discrimination and ever-quickening movement away from the democratic ideal. Ironically, those who are setting the pace are borrowing industrial approaches from the past to bring back aspects of personalized selling that marked the days of peddlers and small stores. Amazon’s price-check stunt proved to be the tipping point that moved physical retailers toward personalization on an industrial scale. The pre-Christmas hijacking, limited though it was, forced brick-and-mortar merchants to recognize that the online and the offline worlds were coming together inside their physical stores. If they didn’t use their aisles themselves to target likely customers with messages driven by the detailed individual shopper information that they had compiled, their competitors would. Once again the UPC, or barcode, would play a major role, this time in the merger of physical and online retailer.

*  *  *

To understand how we arrived at this transition we need to take a closer look at retailing in the late twentieth century—and in particular one retailer that laid the groundwork for change before Amazon. The rise of Walmart is the starting point for exploring how competition, technology, and rhetoric about targeting the customer combined to set the stage for personalization-centered physical retailing. Walmart practically sneaked up as a powerhouse on the American retail industry. “The plain fact is that the market has failed to realize the impact that Wal-mart has made,” said a retail analyst for Smith Barney, Harris Upham & Company in 1989. But, she quickly added, “What it has done has never been done before.”10 From 1977 to 1986 the discount chain increased its annual sales from $678 million to $12 billion, and its net income from $22 million to $450 million. And in 1988 its sales increased 34 percent, to $15.9 billion, and its net income soared 39 percent, to $627.6 million. A New York Times columnist wrote that Walmart is “unarguably the most dramatically successful retail company of the last twenty years. It has leaped out of the South to become the nation’s third-largest retailer after Sears, Roebuck & Company and the K Mart Corporation and is now hurtling ahead with the possibility of overtaking each of them in sales in the next few years.”¹¹

Some speculated that because Walmart had “leaped out of the South” it was not taken seriously despite its surging income. Launched in 1962 by Sam Walton and his brother Bud (the store signs read “Wal-mart” for many years to indicate its founders), the chain systematically radiated “in concentric circles” from its first store in Rogers, Arkansas.¹² Even by 1988 the company operated stores in only twenty-four states, while the corporate base was in Bentonville, Arkansas, far from the traditional halls of retailing power in the Northeast and upper Midwest. Yet during that decade it was already expanding beyond its basic discount store format to include Sam’s Wholesale Clubs and, beginning 1987, adding a grocery section to many of its new, expanded stores, which the company called “Walmart Supercenters.”13 By 1987 Sam Walton, with a net worth that Forbes magazine pegged at $8.5 billion, led the Forbes list of richest Americans by a huge margin.¹⁴

Walton and David Glass, his president and chief operating officer, attributed Walmart’s success to its laser focus on the customer. “As for Wal-mart becoming a household word, we never think about things like that,” Glass told a reporter in 1989. “What we want is for customers in our territories to think of us fondly.”¹⁵ A Morgan Stanley report on the company that year agreed: “Management is insisting that every customer be treated as a guest. . . . This unique attitude is producing a strong response from consumers since service with a smile is often a forgotten commodity in retailing.”¹⁶ This approach democratically included stationing “greeters” by the entrance to the store to help all customers find what they need. The company also tried to draw in customers by emphasizing that the chain was the place where they could “Buy American.” These efforts notwithstanding, most observers agreed that what brought people to Walmart in the first place was its reputation for low prices.¹⁷

That reputation tended to be well deserved, and it created great fear among executives of competing stores. Individual merchants could not purchase items at the kind of scale that gave Walmart leverage to demand rock-bottom prices from manufacturers. But beyond low prices, another reason for Walmart’s success was its absolute concentration on cost control throughout its supply chain. In part, this was simply a matter of survival, as in its early years Walmart could not get large wholesalers to deliver to its rural Southern locations. According to an Associated Press account, the company therefore “had to create its own distribution system, with its own trucks, its own direct dealings with manufacturers, and its own technologies.”18

Sam Walton saw that technology could aid in making that process ultraefficient, and in the mid-1960s he began using computers to track the delivery and sale of inventory in each of his stores. This predated the UPC system, so at first all the items coming in and going out still had to be counted individually. By the early 1980s most packaged goods were labeled with bar codes, and in 1983 Walmart outfitted its warehouses and store delivery areas with the UPC system, including hand scanners so that employees could check in merchandise as it arrived.¹⁹ At the same time, the company replaced standalone cash registers with scanning systems that relayed each individual sale to an on-site computer; the store then reported this information to Bentonville. In 1987 the company greatly speeded the process by setting up the largest corporate satellite system in the world, providing real-time data, voice, and video links to all Walmart stores. Individual sales information from every store was now delivered instantly to the company’s headquarters.²⁰ Walmart made the process even more efficient when it announced that it would no longer buy from wholesalers, instead dealing directly with manufacturers, and in 1988 it proceeded to force these suppliers to link their computer systems with Walmart’s.²¹ As a result of this electronic data interchange (EDI) operation—probably the largest in the United States—paper and fax relationships virtually disappeared.22 Instead, all invoices came directly into Walmart computers for quick processing, and Walmart computers instantaneously alerted suppliers’ computers when an item needed to be replenished.

Writing about Walmart in 1994, a Canadian reporter marveled at the speed and efficiency at which the retail giant was able to restock its shelves. “Wal-Mart’s check-out scanners feed information by satellite dish to the distribution centers, orders are made automatically and products are speeded to a loading dock by conveyors. In the United States, every Wal-Mart store is within a day’s drive of one of about 20 Astrodome-sized distribution centres. The centres are serviced by a fleet of 14,000 trucks, the largest private fleet in the United States.”²³ The reporter described how, using information provided by the company’s computers, Walmart personnel at each distribution center immediately transferred every manufacturer delivery onto the appropriate Walmart truck for carriage to specific stores. No product had to be entered into warehouse inventory, and the speed of transfer “allow[ed] Wal-Mart to re-stock its stores at least twice a week and sometimes every day, compared to the industry norm of every two weeks.”²⁴

These impressive feats, generally recognized as far more accomplished than what Walmart’s competitors could do, were just the tip of an iceberg of efficiency-centered activities Walmart used to wring costs out of products in its now 2,440 stores.²⁵ One ongoing common tactic was to compel suppliers to lower the cost of their products by lowering expenses. Indeed, there were reports that David Glass, who remained CEO until 1993, would even chastise manufacturers for using private jets to travel to Bentonville for meetings with Walmart officials, and for locating their corporate offices in expensive cities. “You don’t tell Wal-Mart your price; Wal-Mart tells you,” said an executive of the American Textile Manufacturing Institute in 1997.26 There were also leaked accounts of Walmart pressuring its vendors to submit to especially low wholesale prices on everything from dresses to kitchen utensils to lawn mowers. Publicly executives were upbeat. “I went [to Bentonville] knowing we were going to get squeezed and wrung and twisted all in positive ways,” the CEO of Liz Claiborne said in 2000 about the Russ clothing brand his firm designed exclusively for Walmart.²⁷ To keep costs extra low, Claiborne employed fewer people for the Walmart project than it would for its department-store accounts, located them outside of New York City to save on rent, and relied more heavily on technology to keep things “faster, quicker, and cheaper when optimized” at the scale of Walmart’s purchases. Even so, Walmart executives persisted in trying to lower the price: “They will ask if that zipper is right, or does that piece of lace trim add value.”²⁸ While Claiborne kept in Walmart’s good graces, Rubbermaid didn’t, and was punished. In the 1990s, the firm was the leading brand-name maker of items such as kitchen trashcans and laundry baskets. “But when the price for the main component in its products, resin, more than tripled between 1994 and 1996,” journalist Leslie Kaufman wrote, “Wal-mart balked at paying increased prices. When Rubbermaid insisted, Walmart relegated the manufacturer to undesirable shelf space and used its market power to promote . . . Sterilite, which made lower-priced nonresin products.”²⁹ Rubbermaid’s profits plummeted and the company was bought by another household goods giant, Newell.30 Newell managed to get Rubbermaid products back on Walmart’s good side.

In the public’s eye, Walmart’s Darwinian fixation on efficiency in the name of democratically low prices was most apparent in its labor practices. Walmart salaries were quite low, employees were nonunion, and huge numbers of their workers didn’t receive health benefits. In regions with a high labor union population, union officials and some government agencies complained that the chain’s size and nonunion practices caused benefits and wages in the area to suffer once Walmart moved in.³¹ By the late 1990s the retailer had abandoned its “Buy American” mantra and instead obtained most of its products from firms that made them outside the United States at much lower cost. Labor advocates complained that these factories were routinely among those with the worst working conditions. Walmart’s CEO in 2000, Lee Scott, adamantly disputed those charges, pointing to a strict code of ethics for overseas suppliers and stating that the company cut off contractors who violated them. “If you are an admirer of capitalism, [Walmart is] the epitome of it,” said economist Carl Steidmann of the PricewaterhouseCoopers consulting firm in 2000. “They are the prime example of the good and bad.”³²

Retailing consultants and Wall Street brokers tended to emphasize the good. By the early 1990s the chain’s efficiency practices reduced its sales costs to 2 to 3 percent below the industry average—savings that translated into billions of dollars in additional revenue for the company.³³ In 1993 AT Kearney management consultant Burt Flickinger noted that “Wal-Mart is the only car on the track with a jet engine under the hood. . . . [It] is so far ahead it’s going to be damn tough for anyone to catch up in this decade.”34 A 1998 article stated that “Wal-Mart’s ability to use data to adjust quickly to market conditions and consumer demand has caught the attention of many top business leaders.”³⁵

Competitor retailers, both large and small, were fearful. Local merchants in small towns where Walmart had pitched its huge tent complained that they could never get manufacturers to sell them goods at near the wholesale prices that Walmart, with its formidable buying leverage, paid. Even the huge Kmart, Sears, J. C. Penney, and Target chains, Walmart’s most direct national competitors in the broad discounting realm, struggled mightily to keep up with the market share it had gained at their expense.³⁶ Department stores were especially vulnerable to competition during the 1990s, when Walmart was spreading more broadly than ever across the United States, and many analysts saw the department-store form as inherently weak. The New York Times summarized the view held by many: “Once a growth industry that racked up profits from a rising population tide,” it wrote, “retailing has matured into a business where one store’s gains now come only at the expense of another. The massive overbuilding of shopping complexes during the 1970s has left America ‘over-malled.’ Yet Americans’ appetite for these products has slackened. The baby boom generation that turned shopping into a leisure activity is aging, and becoming more interested in saving for the future.” Many also thought that traditional department stores, such as Macy’s, J. C. Penney, and Burdines, were caught between discounters such as Walmart and specialty stores such as Gap, Ann Taylor, and Crate and Barrel—and even between manufacturers such as Ralph Lauren, Wedgewood, and Burberry that sold their products in department stores. Additional niche competition came from twenty-four-hour ordering departments of catalog companies such as L.L. Bean and Sharper Image. “Department stores are at the end of their life cycle and have to be reinvented,” the head of an investment bank told the New York Times.37

A number of Wall Street takeover specialists exacerbated the situation in the 1980s. Confident they could sell department stores’ real estate and unrelated businesses for a sizable profit, this group engineered a succession of department-store buyouts that caused the debt of the new owners to skyrocket, with retail margins that could barely support it. Those huge repayment requirements, noted Investor’s Business Daily, “hurt [the companies’] ability to upgrade stores and systems to better compete with discounters like Wal-Mart Stores, Inc.”³⁸ The major economic recession in 1990 and 1991 made repayment especially onerous, and the fallout was devastating. From 1990 to 1995 a total of 316 department stores failed, and two such companies, Federated Department Stores and Macy’s, entered Chapter 11 bankruptcy protection.³⁹ A healthier Federated bought Macy’s once it emerged from Chapter 11 in 1994. Federated had under its umbrella some of America’s most famous regional store brands, from Bullock’s in the West to Lazarus in the Midwest to Abraham and Strauss and Bloomingdale’s in the East. Many hoped that Federated CEO Alan Questron, an efficiency fanatic, would set an example by steeling his merchants to face Walmart and other discounters.⁴⁰

The supermarket business likewise found it difficult to keep pace with Walmart. A spate of acquisitions during the merger-happy 1980s–early 1990s brought many regional brands under a few prominent names—Kroger, Albertson’s, Safeway—while some independent retailers—Harris Teeter, Publix, and Wegmans, for example—held their own as respected forces within the industry.41 But they all worried about Walmart as its Sam’s Club and Supercenters with grocery sections started to open across the United States. The discounter’s grocery prices were often set artificially low, as executives anticipated that the store would earn higher margins when grocery customers went on to also shop in the nongrocery aisles. Walmart didn’t follow supermarket-industry tradition whereby low margins were made up via a second revenue stream from manufacturers that paid for special placement and promotion of their goods. Grocers historically would charge manufacturers for the right to place certain goods on the shelves (“slotting allowances”);⁴² take some of the money manufacturers pay for promoting particular products and use it for other purposes (“diverting”);⁴³ and take advantage of promotional pricing by ordering a sale product at a reduced price, then holding it until the sale is over and making higher margins on the regular price (“forward buying”).⁴⁴ Because manufacturers didn’t have to play these sorts of games with Walmart or Sam’s Club, they were happy to give Bentonville a better deal.⁴⁵ A historical review appearing in 2012 in the trade magazine Progressive Grocer noted that the supermarket industry in the 1990s “was put on abrupt notice by the ‘Bentonville Behemoth’ . . . [t]hat the rules of the game were in the process of being permanently upended by its efficient, and quite literal, march into the food business.” That food business, the magazine added, “found [Walmart] at once becoming the most feared, loathed, and respected of any retailer on the planet, before or since. Indeed, the weakest links in the industry—those that ignored or dismissed Walmart’s potential impact on their marketing turf or just refused to evolve to meet new standards—rapidly fell by the wayside in a spate of acquisitions, closures and sell-offs.”46

Department stores and supermarkets developed a variety of strategies to compete in a Walmart world. Some retailers supported community efforts in their towns to oppose the establishment of the “Bentonville Behemoth.” These fights often focused on academic studies and anecdotal evidence concluding that the arrival of a Walmart drove local merchants out of business and devastated downtown areas. Typically these battles pitted citizens who wanted to retain their traditional shopping districts against those who supported the potential economic benefits, such as new construction jobs and the influx of shoppers from outlying areas. Although traditional shopping did win at times, it was an uphill battle. Some local retailers pursued their own programs in the hope of matching Walmart’s vaunted efficiency levels: by the early 1990s, the use of barcodes, electronic data interchange, and the so-called just-in-time replenishment of stock was becoming common, especially in groceries.⁴⁷ The department-store retailer Nordstrom, for example, touted its rollout in 1992 of an inventory system that used radio-enabled UPC scanning devices and touch-screen personal computers “to help merchandisers make smarter buying decisions and allow them to monitor the unexplained loss of goods.”⁴⁸ Around the same time, Federated was forging ahead with software aimed at making planning and buying more efficient.⁴⁹ These advances could help increase the rate of inventory turnover and reduce the need for markdowns and sales. “That’s the area of greatest leverage,” agreed Stewart Neill, Saks Fifth Avenue’s vice president of management information systems. “If you can speed up turn[over]s . . . that will increase profit more than anything else you can do.”50

By the late 1990s, according to Investor’s Business Daily, department stores were turning the corner in efficiency, noting in an article that through the decade they “had been building muscle by consolidating.” The article added that “to cut costs and better serve customers they’ve built clout with vendors in the same way.”⁵¹ Nevertheless, when department stores and supermarkets competed toe to toe with Walmart for better margins, they lost. Even its giant competitors Kmart and Sears couldn’t match Walmart.⁵² These big chains poured millions of dollars into the kinds of satellite systems and back-end technologies Bentonville was using, and yet they still experienced smaller margins, decreasing customer counts, and lower profits. And while Walmart was often first and foremost in the minds of many department-store and supermarket executives during the 1990s and early 2000s, this group also worried about direct and indirect competitors in their own industry. For department stores this meant big discounters and specialty stores, while supermarkets saw Kmart (which carried groceries), limited-inventory discount stores (such as Dollar Store), deep-discount drugstores (such as Drug Emporium), and convenience stores (such as 7–11) as stealing shopper time from their aisles.

In the face of both the efficiency treadmill and considerable competition, supermarket executives began to argue that, beyond the need for back-end productivity, the only way to become profitable again was to find ways to differentiate themselves from the discounters. Advertising Age captured the challenge in a 1993 article noting that “much of the supermarket industry seems stuck in the same retailing midway that’s creating troubles for old-line retailers like Sears, Roebuck and Co. and local department stores.” It quoted an advertising executive who encouraged executives to shift from trying to fill customer desires democratically to following the pragmatic need for discrimination. “No retailer can be all things to all people,” he argued. “If you don’t have a clearly defined niche in the marketplace, and you don’t generate top-of-mind awareness at least among that segment of the market that you say you stand for, you are lost.”53

Sure enough, retailing discrimination started to catch on. Bill Bishop, founder of the Willard Bishop packaged-goods consultancy, remembers urging skeptical CEO Danny Wegman of Wegmans Food Markets to adopt the barcode system for collecting information on his customers. “When we first started working in the ’80s on loyalty data, people like Danny Wegman didn’t want to treat people differently,” Bishop said. But, Bishop said, Wegman ultimately came around as he began to understand the value of identifying the 20 percent of shoppers who brought in 80 percent of the revenue.⁵⁴ The inevitability of discrimination was aided by an existing retail business truism stating that the declining rate of U.S. population growth meant competition for new customers would therefore become substantially more rabid and expensive than in the past. Many in the industry argued that efficiency meant retailers needed to place more emphasis on retaining good customers than on finding new ones—and that the way to do that was to learn as much as possible about those good customers in order to know what persuasive levers would keep them returning.⁵⁵

Most retailing experts also agreed that competing on price alone was not a winning strategy against Bentonville or other discounters. They believed that Walmart was succeeding also because in shoppers’ minds its stores were linked to solicitous service. True, pricing had to be low, the experts acknowledged, and to compete successfully certain high-profile items needed to be heavily discounted. But they also cautioned that relying on the lowest price to differentiate one store from its competitors would only encourage a deadly race to the bottom against another retailer that might be better able to sustain losses. Instead, according to a 2003 essay in Progressive Grocer, the key was finding a large enough niche in the marketplace that eluded Walmart and that served specific, carefully designated customers and their interests. “Being where Wal-Mart isn’t,” the article noted, “means targeting customers not served well by Wal-Mart, addressing the needs of that segment in a compelling value proposition, and then designing value innovations that are hard to copy or disadvantageous for Wal-Mart to duplicate. And, of course, the exciting outcome . . . is maximized sales, more loyal customers, and an impenetrable barrier to the encroaching mega-chain.”56

At its core, this idea wasn’t new. Traditionally department stores and supermarkets had attempted to present a particular image to draw certain types of shoppers to them over their competition. In the early 1990s, though, many worried that retailing personas were not sharp enough for the new competitive era. Department stores in particular were singled out: mergers clouded the original brand’s personality; store brands were downplayed in favor of nationally advertised merchandise (such as Polo, Etienne Aigner) carried by various merchants; baby boomers didn’t share their parents’ loyalty to department stores; and department stores tended to follow discounters in focusing on price-oriented rather than image advertising. The lack of a clear-cut image pushed many shoppers into the hands of Walmart, with its constant drumbeat of low prices and service. “Retailers need to prepare to fight harder for their customers,” a retailing consultant told Advertising Age in 1988.57 A senior sales executive at Chicago’s Marshall Field’s department store agreed. “There was a time when we took our existing customers for granted and saw chasing other stores’ customers as the key to growth,” he said. “Now all that’s changed, and we see our future growth coming from our current customer base.”⁵⁸

Supermarkets had done at least a little better in establishing individual images. In 1986 a stock analyst told Advertising Age that the Albertson’s supermarket chain’s “niche has been to offer attractive stores, specialty departments, nonfood merchandise, and a little more service at a slightly higher price.”⁵⁹ With few exceptions, though, supermarket images were not immune to encroachment from Walmart, and stand-alone grocery businesses were losing ground to the discounters that also offered food aisles. This was a new world.

A refurbished loyalty program seemed an obvious solution. But instead of trading stamps, many stores took a cue from airline and credit card companies and offered reward programs, which for retailers meant awarding points based on purchases so that customers could redeem them in the form of discounts and rebates at checkout. “It’s hard to walk into a supermarket or department store these days without being asked to join their ‘preferred customer’ club,” noted a 1993 Advertising Age article.⁶⁰ Naysayers saw such activities as misguided. A director at the Bain & Company consultant firm considered it a way of “bribing” customers to continue buying, a strategy that he said rarely pays off in the long run.61 Don Schultz of Northwestern University’s Integrated Marketing program agreed. “I see an awful lot of [marketers] who get into it, can’t figure out how to make it work and get out pretty quickly,” he told Advertising Age. Walmart had no loyalty program.

Some urged that loyalty programs be transformed rather than discarded. The real utility in such programs, they insisted, was not in bribing customers but in the valuable insights they could glean about them.⁶² Walmart’s target audience consisted of shoppers from one socioeconomic background, so competitors that collected data via rewards programs could identify other types of customers and pursue those groups. Shoppers who saved money as a result of their membership in the program would be encouraged to continue buying from the same retailer, while at the same time the retailer would learn much about these repeat customers. Knowing the buying habits of the most frequent and highest spending customers could enable a store to develop its “best customer” profile and therefore its most effective market niche. The same technologies that enabled stores to collect this information also enabled them to reach out to customers with individualized rewards or discounts. And of course other stores wouldn’t even be aware of—and so couldn’t match—these specific offers. An executive from the Epsilon database company referred to such offers as “a stealth weapon in frequency marketing.”⁶³

Catalog companies likewise began adopting a targeted approach to reach specific customers. By the late twentieth century these companies faced huge costs associated with producing and mailing their promotional materials. A data analyst for the catalog company Lands’ End noted that “we’d go bankrupt quickly” without sophisticated analysis to cull the most likely buyers of items in specific catalogs from the twenty million names in his firm’s database.64 The high cost of producing and mailing catalogs may explain why retailing experts have tended to exhort retailers to find ways to reward the “most valuable (and most profitable)” customers in the store itself.⁶⁵

Department stores were slow to adopt a personalized approach. In the early 1990s they were preoccupied with trying to make their behind-the-scenes operations more efficient as the best way of keeping their books balanced. For them the challenge was to be profitable by offering their target clientele the right merchandise when they wanted it and in the right amounts and sizes, and at the right prices. Still, many department stores did attempt to take what data they had and made targeted mailings based on gender, income, and lifestyle. Taking a page from their forebears, several built personal-shopper programs aimed primarily at corporate customers short on time or fashion sense.⁶⁶ Pursuing technologies for in-store personalized deals based on rewards program data didn’t gain traction for at least another decade.

In contrast, the idea did begin to take hold in the supermarket business, as those executives were recognizing that individualized shopper attention could prevent defections to competitors. “The average shopper is spending $50, $60, $100 a week in a supermarket and yet the store is not in personal contact with that customer,” Progressive Grocer’s editor told the New York Times in 1991.67 Not only did supermarket executives recognize that personalization could establish a level of connection with customers that would keep them from shopping elsewhere, they also believed that their suppliers would pay them to do it. A power shift had occurred in the grocery business, as manufacturers now depended more than ever on stores for reaching people with messages about their products. One reason for the change was the shopping environment; research conducted in the late 1980s and into the 1990s consistently showed that a large percentage—some argued as high as 80 percent—of shoppers made their decisions on specific brand purchases in the supermarket aisles.⁶⁸ Manufacturers saw this as an opportunity to use advertising, and especially coupons, more effectively.⁶⁹ Procter & Gamble (P&G) CEO A. G. Laffley referred to it as “the first moment of truth,” the instance when all the money the company had invested to push its creations was tested.⁷⁰

Traditionally, P&G and other manufacturers of branded products spent greatly on advertising their products in mass media. They also placed discount coupons in circulars that were distributed by mail, in stores, and in newspapers. Americans redeemed roughly 1.7 billion coupons in 1993.⁷¹ Considerable cost was involved to produce and redeem the coupons (each one had to be hand-counted), and the brand manufacturers tended to see them as an unfortunate cultural addiction that they wished they could eliminate.⁷² Unable to do away with them outright (though P&G did experiment with this),⁷³ they tried other tactics to increase their effectiveness. One involved placing UPC codes on coupons for scanning as a way to reduce their processing costs.⁷⁴ Companies also began to focus more on increased in-store distribution, which they found could generate four or five times the response rate of newspaper inserts.

Although supermarket operators profited by redeeming shoppers’ coupons, they realized they could earn more by allowing vendors to set up advertising or coupon-distribution technologies in their stores and then charging them for a portion of their earnings—typically either a flat fee or a guaranteed percentage (sometimes as much as 25 percent). Advertising Age estimated that large chains such as Kroger, which operated 1,234 supermarkets in twenty-four states in 1990, could earn $1 million or more by allowing the dispensing of coupons alone. “This revenue-generating potential doesn’t go unnoticed in an industry that operates on a tight profit margin,” the trade magazine noted.75 A horde of technology companies then descended on supermarkets, with various mechanisms for in-store advertising and coupon circulation. Depending on the store and the region, shoppers might see electronic advertising signs, various video information displays, shopping calculators (with ads), ad-supported radio networks, and both shelf-based and checkout-based coupon dispensers. “It’s becoming a blur,” said an ad agency executive about the profusion of promotional options.⁷⁶

Progressive Grocer estimated in 1991 that 12 percent of the seventeen thousand chain stores and 3 percent of independent grocers offered some sort of electronic coupon program at checkout. Many tried tailoring those coupons to the purchaser’s buying habits. Von’s supermarket chain of southern California, for example, replaced what Progressive Grocer called a “standard” loyalty program in 1990 with one that cross-indexed “by customer and product category” the purchases of the program’s three million members. Based on this information, Von’s mailed specific promotions to members, matching what they had previously bought with participating marketers’ coupons.77 King’s Super Markets sent targeted letters to its reward card holders. “It’s possible to do a mailing to kosher customers,” said an executive. “It’s possible to write to people who never buy pasta to try to win them over.” The store even mailed letters to customers who hadn’t used their cards in a while in an attempt to rekindle their interest.⁷⁸

By today’s standards, these tailored actions seem a fairly simple targeting of customers by segments as opposed to the more current sorts of multivariate analysis that yield very specific insights into shopping habits. Nevertheless, the scanners spewed out so many data points that the industry found itself overwhelmed. The New York Times noted in 1991 that “the data is so hard to use because its volume is intimidating and using data of any kind in the supermarket industry is new. The very fact that stores can find out almost anything they wish from the data is bewildering to some managers. Should they pursue lost customers? Senior citizens? Young couples with more than one child? Once they do choose a group, what enticement should they offer?”⁷⁹ The manager of Ukrop’s Super Markets in Richmond, Virginia, agreed. “You have to be careful because there’s so much [scanner data] out there you can be run over by it.”⁸⁰

Despite the immense analytical challenges, many supermarket executives believed that they couldn’t afford to ignore the personalizing possibilities of database marketing in view of the intense competition among stores. Yet even though the biggest suppliers had the cash and expertise to conduct these database explorations, the grocers typically weren’t willing to share their data with them. An exception was a firm called Catalina Marketing, which was enlisted by grocers to connect purchased items to customers only by their loyalty numbers, not by any personal information. Based on Catalina’s straightforward analysis of an individual shopper’s purchases over time, discount coupons would be printed at checkout using a small Catalina-installed system; manufacturers paid the firm to generate the coupons.81 So, for example, a shopper who buys shaving cream every six weeks might, five weeks after his last purchase, be presented with a coupon from a shaving-cream brand he didn’t use—or from one he did use, depending on the brand that paid for the coupon.⁸² In 1994, Catalina was sharing its coupon revenues with seven thousand supermarkets.⁸³

Many of the merchants that didn’t use Catalina were simply overwhelmed by the amount of data involved or were generally reluctant to adopt personalization. One such grocer, Von’s, found it so difficult to conduct sophisticated analysis of its database that the grocer fell back on distributing electronic coupons based just on the presentation of a frequent shopper card.⁸⁴ A year later, however, Von’s was back again with a rewards program that tried to match shopper purchases with manufacturers’ coupons.⁸⁵ Because of the consistently low level of statistical capability among grocers, some in the industry believed that supermarkets would ultimately have to bring in sophisticated third parties—even major high-tech firms such as IBM and NCR—to cull purchase information and buy third-party data to encourage an actual “two-way dialogue between the retailer and the customer [in the form of] personalized communication.”⁸⁶

By the mid-1990s the level of personalized communication supermarkets conducted with their customers didn’t seem that much different from department stores’ approaches to their clientele via in-store frequent-shopper point accumulations and targeted mailings. The insulated nature of both industries, their lack of data-analysis capabilities, and the high monetary and competitive costs involved all had stood in the way. But despite the industries’ reluctance to embrace the new technologies, they soon concluded that database-driven personalization was a valuable activity that should become an integral part of retailers’ relationships with shoppers—if not immediately, then down the road. And as it turned out, developments in electronic commerce were making that road more navigable. Yet the department-store and supermarket industries still needed another fifteen years to make widespread in-store personalization a priority.

The stirrings of electronic commerce date back to the 1970s, and the 1980s saw the creation of a number of key endeavors, including France Telecom’s Minitel online ordering system (1982) and CompuServe’s Electronic Mall, the first comprehensive electronic commerce service in the United States and Canada (1984). The 1990s ushered in transformative changes that ignited e-commerce along the lines we see today—particularly at mid-decade, when developments reinforced retailers’ growing interest in using technologies for personalizing their relationships with shoppers.87 A signal of things to come was a 1993 Adweek article urging the advertising industry to “take note” of the internet. The emerging technology has, the essay said, “the potential to become the next great mass/personal medium.”⁸⁸ The U.S. government had recently begun to allow private activities on the internet, which had previously been a vehicle mostly for university and military researchers. The major internet-related breakthrough for advertising and commerce was Tim Berners-Lee’s 1989 invention of the World Wide Web system of hyperlinks. It led to the creation of graphical Web viewers (browsers), notably (in 1993) Mosaic and, a year later, Netscape, which were the first popular ones. Browsers allowed people to peruse photos and information regarding goods for sale on websites, and to make purchases of those goods directly on the sites.

Mosaic was an apt name for the browser, as websites are typically a mosaic resulting from “a two-way interaction between users and a company’s webserver (the computers that host internet content).”89 When a user types a Web address into a browser, that person’s computer is requesting to view the content of a specific page. Although a website’s words and images seem to appear almost instantaneously, the conversation between a user’s computer and the website’s server covers several complicated steps. In general, each item on a page is requested and issued individually. Although the website dynamically generates content, the page itself is “assembled” on the user’s computer, which downloads software code from the website and executes it on the user’s system in order to create the Web page.⁹⁰

The origin of the content you see on the browser depends on the site that you requested. Sometimes most of the content comes directly from the website you chose—called “first party”—and whose URL appears at the top of the browser. But often at least part of the content is delivered from elsewhere—“third parties,” which deliver material to your chosen site typically with the permission of that site. Commercial messages are the most common third-party additions. Of particular interest to advertisers is the Web’s ability to identify certain aspects of the person receiving that message even without the person realizing it. When you go to a website this interaction is recorded on the website’s server, along with various information regarding the interaction, such as your computer’s internet protocol (IP) address—a string of numbers associated with your computer. The IP address can reveal your computer’s geographic location within a few hundred feet. The website can also record the browser you are using, the language you are using to communicate, and the speed of your transmission. This information enables the website to tailor commercial content appropriate to your device (that is, the correct language and the right graphical layouts).

But this sort of information, helpful though it is, isn’t nearly enough for advertisers hungry for audience information. For one thing, IP addresses aren’t useful for identifying a specific user indefinitely. Internet service providers might change home users’ addresses every few days, and when multiple computers share one IP address, as is the case with most standard home networks, it’s impossible to know whether one computer or several computers are being tracked. One way to be sure of a user’s identity was to ask individuals to log in to a website each time, though early on this was not considered a particularly good alternative. Another issue involved tracking purchases on a site. At first browsers were set up so that people wanting to buy more than one item had to buy each item separately; Mosaic had no way to store choices—the electronic shopping cart had yet to be instituted—and so a person could not purchase the items together.

A crucial step in the ability of advertisers to collect information on individuals was the invention of the computer cookie in 1993. Cookies are typically small text files that are placed on a user’s hard drive by the website the individual is visiting, and only the person or organization placing the cookie can access it. Even after a user exits out of a website, the cookies placed by the website remain on the individual’s computer. Companies quickly realized that cookies could enable them to gather much information about an individual, including the dates and times when the person visits their website. Moreover, if a website could get a shopper to log in or otherwise provide the person’s name, email address, or other personal information, this data could also be added to the cookie. The website could retain the information on its servers even if the user erased the cookie (as privacy activists increasingly advocated), enabling it to reconnect the information the next time the person logged in.

Advertising firms also discovered that they could recognize people across sites by placing “third-party” cookies on the user’s computer, enabling the marketers to follow individuals from website to website—any site from which the marketers had permission to place cookies—and to record a person’s presence and activities on each site. An advertiser, a network of advertisers, or a database company that had made cookie-placing agreements on thousands of websites (or more) could compile a substantial listing regarding websites an individual had visited—and make inferences about the person’s lifestyle and interests and how those might affect the individual’s purchase patterns.

The cookie was the most crucial of a range of emerging developments that deepened the notion that the Web was a place for promoting products as well as collecting data on individuals and then using that information to entice them to make a purchase.91 But despite exhortations from trade journals, retailers approached tailored communication rather irregularly and experimentally, even as the Web quickly became a model for compiling data on consumers. “You don’t want your company to be left at the starting gate as more nimble competitors race toward the pot of gold in electronic commerce,” lectured a Computerworld columnist.92 An article in Progressive Grocer predicted that “everything we currently imagine we know about the consumers will be redefined in the next 10 to 15 years.” The article urged readers to “just think of the [small] differences between micromarketing to physical neighborhood and a cyber community. . . . Or consider what service means in a virtual shopping environment in which the retailer never physically meets the consumer, but where they know more about that shopper as an individual than any ‘place-based’ retailer who sees the consumer three times a week.”⁹³

But exactly when this complete transformation was going to occur was open to question. The Computerworld article, titled “Old Rules Apply in the Marketplace of the Future,” cautioned that retailers must create an electronic marketplace “that supports the entire procurement process” and is not just caught up in the enthusiasm of selling as long as the transactions are secure.⁹⁴ (Visa, MasterCard, Microsoft, and Netscape were already hard at work on transaction security.) The Progressive Grocer envisioned a time frame far beyond the typical five-year business plan. An editorial by the retailing editor of Computerworld in 1996 underscored that Web commerce was a long-term wager. “Current betting is that Web-based shopping will be slow to catch on because it is only appropriate for certain types of purchases,” he wrote. He suggested that “anything that requires real-time, hands-on examination is probably safe—for now.” He predicted that new multimedia development tools, such as Sun’s Java, could eventually deliver “Jetsonesque” experiences such as online digital dressing rooms and digital agents to handle “time-consuming” personal shopping. But when these sorts of technologies do become reality, he noted, “like it or not, you’ll have to make sure your web presence is second to none.”95

The efforts of the trade press notwithstanding, many physical retailers were hardly persuaded that the Web deserved their attention, though some were experimenting with an online presence. In anticipation of the holiday shopping season in 1996 some major retailers launched websites, including Sharper Image, Saks Fifth Avenue, F.A.O. Schwartz, Eddie Bauer, J.C. Penney, and Omaha Steaks. Private online services such as America Online and CompuServe also offered electronic stores from companies with physical counterparts, but, according to a New York Times article, the overall internet “has a far greater variety.” One way a retailer might achieve a Web presence at that time was to join a so-called cybermall, which presented various stores under the umbrella of one master site. In 1996 Saks, Eddie Bauer, and Omaha Steaks were part of a cybermall owned and operated by Time Warner Cable. Such approaches were more prevalent in the pre-Google world, when Web directories such as Yahoo! and search engines such as Infoseek and Excite had yet to optimize their navigation functions. “Finding what you are looking for can be both difficult and time consuming,” the Times article cautioned, “and many newcomers to the Web will tire of having to jump from site to site and item to item, an experience that has earned Net shopping the reputation of ‘Death by 1,000 clicks.’ Others will enjoy the adventure and find it like window shopping, without the sore feet.”96 A year later, an article in the Washington Post also reflected online retailers’ desire to be seen as “growing and quickly becoming more mainstream.”⁹⁷

With an increasing number of people using the internet for shopping, the case for physical stores establishing a Web presence seemed apparent. At this time direct-marketing advertisers were setting the pace online, serving ads to individuals based on increasingly specific data pertaining to their online habits as well as their offline habits. (Offline information was purchased from non-Web database firms such as Acxiom.) Meanwhile, online-only retailers such as Amazon were using cookies to show returning shoppers items they might be interested in buying based on their previous activity on the site, as well as on the retailers’ data analysis of other people they deemed to have similar shopping inclinations. Some observers began to suspect that online retailers were also using their databases to tailor prices to individuals, though the merchants weren’t saying and direct evidence for this was slim. Data collection for commercial purposes increasingly ignited discussions in the press, academia, the federal government, and state governments about privacy and surveillance and their potential limits, and about whether any limits should be voluntary or mandated by law. Web enthusiasts played down any potential harm to consumers and instead touted the benefits of offering ads and products of specific relevance to an individual. They also argued that data-driven Web commerce was a great engine of American economic growth.

Physical retailers observed these developments and recognized that data-driven targeting and personalization held great promise. They increasingly parsed their customer and loyalty data for targeted mailings of sales and catalogs. Yet they were reluctant to spend the tens of millions of dollars it would to take to mount “a web presence second to none,” concerned that the Web was actually geared toward catalog rather than brick-and-mortar sellers. Indeed, it was no coincidence that several of the stores featured in the 1996 New York Times article had strong mail-order businesses. Firms such as Eddie Bauer and Omaha Steaks were already well positioned for handling online ordering and fulfillment; firms that existed outside the mall, such as Lands’ End, were likewise good candidates for the Web. Walmart became involved in 1996 with the help of Microsoft and then revamped the site on its own several years later. Department stores, however, typically were not prepared to compete. Bloomingdale’s tried establishing a website, but by 1999 it no longer existed.98 As of 2004 Lord & Taylor still didn’t have one at all. Macy’s offered a relatively thin site at first, but in 1999 its parent company, Federated Department Stores, expanded it and also purchased catalog retailer Fingerhut at a cost of $1.7 billion to hone its logistical and fulfillment expertise.⁹⁹ A Federated executive noted at the time, “None of us knows how big e-commerce will be. But I’m convinced that the company that gets the most out of commerce will have to have database marketing and order fulfillment capabilities—not just a fancy website.”¹⁰⁰

Although mindful of the direct-marketing model, Macy’s executives and their counterparts at other department stores during the early 2000s typically viewed internet commerce as a business to be run parallel to, but separately from, their physical store aisles. The databases for each were typically unconnected. High-end retailer Bloomingdale’s, for example, used a sophisticated system called Klondike to identify and cultivate its top customers who shopped either in the physical store or by calling its toll-free ordering number, but this system was not connected to activities on its website. As early as 1998 Walmart began to allow customers to pick up Web orders in a store near their home.101 But this policy did not result from enthusiasm for the new retailing form; rather it grew out of a specific need—many customers did not have a credit card and so needed to pay in person with cash. Beginning in 1998 Macy’s also allowed store pickup for online orders. In 1998 it brought its website under a new interactive division based in San Francisco. The division’s head said that the website had the potential for increasing store sales by attracting younger customers and encouraging purchases from home, especially in markets without Macy’s brick-and-mortar stores. Yet the chain’s leadership was keen for shoppers to see Macy’s.com as separate from the physical establishment so that the former wouldn’t be “eating away at its department store trade,” reported the Atlanta Journal and Constitution. “Macys.com is designed to attract customers who don’t like department stores.”¹⁰²

In the grocery industry various attempts had been made to sell food online from almost the Web’s commercial beginnings. Services such as Webvan and Peapod received much attention for attempting to carve out a niche that involved quick delivery of perishables that had been ordered online. Most lost money (as did many Web ventures at first, including Amazon) and were miniscule in reach. Larger, mainstream grocers such as Kroger and Albertson’s experimented with Web-driven home distribution services in the 1990s and early 2000s, but they saw the internet’s utility mainly as an extension of their promotion business. In fact, supermarket executives saw the Web as even more distinct from their physical stores than department stores did. A 1998 Progressive Grocer article encouraged the personalization possibilities associated with promotional use of the internet. “In on-line shopping models,” the article noted, “on-line ads, coupons, new product announcements and a wide range of promotional events can be delivered precisely to specific households with specific demographics and buying histories.”103 Clearly the public was ready to take advantage of ways that the internet could benefit their grocery shopping. The New York Times noted that the number of shoppers who visited the coupon-dedicated site Supermarkets.com every month grew from twenty thousand in early 1998 to three million just over a year later. Coupon sites took off when manufacturers realized that shoppers would not go to their corporate websites in search of coupons. They therefore made deals with many sites, including supermarket sites, to circulate the deals. Consumers could print coupons and redeem them at their grocery store. Although traditional newspaper inserts and postal circulars overwhelmingly remained the predominant means for receiving coupons, the Times noted that “on-line coupons offer a major advantage: customization. Instead of picking through a random selection of coupons that arrive in the mail or tucked into the daily newspaper, shoppers can fill out on-line questionnaires specifying their interests, ZIP codes and shopping habits and then receive E-mail with relevant offers on a regular basis.” Industry analyst James McQuivey predicted that the targeting and personalization of the online coupon business would evolve rapidly. “This is just the first step in the evolution of on-line coupons because there’s an enormous opportunity here,” he told the Times. “In the next two to three years, companies will figure out how to get coupons more and more easily to consumers.”104

Electronic coupon delivery proved to be a growing part of the armamentarium that supermarket chains used to try and keep customers from straying to other grocery-only competitors or the food-and-everything-else stores of Walmart and other discounters. However, the grocers realized they needed to associate coupon distribution with their own identities as opposed to using freestanding coupon sites. One way to do that would be to present their database-determined valued customers with targeted offers. A variety of emerging firms were engaged to do heavy lifting. The Food Lion chain hired a company called Bigfoot Interactive to provide email communications solutions “for ongoing customer retention and coupon incentive efforts.”¹⁰⁵ Foodtown hired S&H Greenpoints, the twenty-first-century incarnation of the popular Sperry & Hutchinson Green Stamps, to help manage its rewards program for points “that can be earned online, offline or via credit card, and can be redeemed at both clicks-and-mortar and bricks-and-mortar stores.”¹⁰⁶ And the Great Atlantic & Pacific Tea Co., still in business but far smaller than in its heyday before World War II, analyzed the A&P loyalty program’s database and created targeted coupon offers for frequent shoppers. The program, called One-to-One Direct, sent the offers via the U.S. Postal Service.¹⁰⁷

Inside supermarkets themselves the grocers launched a new wave of targeted-coupon distribution. Prior to 2000, they typically distributed personalized in-store coupons to customers as they exited, most often via Catalina printers. The redemption rate of Catalina coupons was higher than that of newspaper coupons—6–7 percent as opposed to 3 percent. Still, grocers and manufacturers complained that shoppers lost many of the checkout coupons before they returned to shop again. While not giving up on the Catalina printers some grocers, in response to the growth of general coupon websites, started to target shoppers in the early 2000s as they entered the store or moved through the aisles, mostly through newly installed kiosks. At the kiosks shoppers could swipe their rewards cards in machines that then generated coupons tailored to what the grocer and participating manufacturers believed would interest the customers as they moved through the store. A more sophisticated technology involved radio-connected carts or handheld devices that were linked wirelessly to a store computer and that offered shoppers various supermarket services (such as notifying them when deli orders were ready) as well discounts based on database information; in 2004, both Albertson’s and Stop & Shop were experimenting with these systems.

Neither the kiosks nor the other high-tech gizmos caught on. Perhaps they were simply too expensive for stores to justify their use, and maybe customers simply found them too complicated to operate. And possibly grocers were growing a bit impatient with rewards programs overall, which too often were compromised by normal human behaviors. “Quite frankly, we’ve been hard-pressed to find retailers that are doing the customer loyalty program with specific customer successes,” noted a retailing consultant in 2007. “Shoppers forget to bring their cards for every trip, and/or borrow someone else’s, and the integrity of the data is then compromised. Pretty soon all the intelligence that an expensive loyalty program has gathered is useless.”108 Still, the logic linking Web promotion to the physical store was clear: if Web marketers could track people’s activities and send them coupons online, perhaps grocers could apply the same process even more successfully as people moved through the stores. And for manufacturers anxious to identify likely and valuable customers and to learn as much about them as possible, the ability to personalize a deal as a person faced an actual “first moment of truth” could be an exciting development.

The mobile phone greatly simplified efforts to achieve this goal. For the first time the shopper, rather than the merchant, brought the connecting technology into the store—and that technology could be used to reliably identify the individual. An early approach, the Mobile Rewards system from Boston-based MobileLime, transformed a shopper’s cell phone number into a unique identifier. “Employing the cellular-based payment and loyalty capabilities, the single-store operator will offer shoppers instant, individualized rewards without the cost and inconvenience of standard in-store reward cards,” noted Progressive Grocer in 2005. Mobile Rewards enabled a supermarket “to tailor special promotions and send real-time offers to customers” as they moved through the store; shoppers were sent text messages and were invited to use their cell phone to call a dedicated toll-free number for bargains.109 In a follow-up piece two years later, the trade magazine concluded that the program was transforming the “lowly mobile phone into a marketing, loyalty, and payment device.”¹¹⁰ In the United States at this time, the plain flip phone (or “feature phone”) was by far the most popular type of cell phone among the 77 percent of adults who owned a mobile phone.111 Smartphones—devices that connected to the internet as well as to cellular phone networks—represented just 4 percent of the mobile market.¹¹² Those numbers would soon change radically with Apple’s release that same year of the iPhone, which reshaped Americans’ understanding of mobility and influenced the creation of the open-source Android smartphone operating system that would ultimately dominate the market.

By 2009 only about 29 percent of Americans owned smartphones, but those who did tended to be upscale and were therefore desirable to marketers.¹¹³ Clearly if stores could use the flip phone to identify shoppers and cultivate personalized relationships with them, they could do the same thing with the smartphone, but with far more gusto: these phones could accept cookies from websites, and they could be tracked. At this time most of retailing wasn’t yet taking advantage of the smartphone’s capabilities, but industry analysts believed that the devices, especially Apple’s version, indicated a need to rethink how stores approached shoppers. “Few devices have revolutionized the way people communicate and live more quickly than the Apple iPhone,” marveled a Progressive Grocer columnist in late 2009. With the advent of home computers and now cell phones, many analysts believed that shoppers for the first time had supreme leverage over the retailer because they could access these technologies to obtain coupons, investigate product quality, compare prices, and make purchases. “The consumer is now in control—think ‘Command Center,’ noted the Progressive Grocer.¹¹⁴ “The challenge for retailers,” the writer emphasized, “is to figure out how to use the technology, and your relationships with shoppers, to the benefit of your [store’s] brand.”¹¹⁵

Supermarkets still considered Walmart to be their main nemesis, but their concern had moderated somewhat when the chain halted its rabid march across the United States in favor of overseas expansion. Some in the industry also pointed out that Bentonville wasn’t bulletproof even on a supermarket’s home turf. Because the discounter’s primary customers were from the lower economic end of the earnings spectrum—albeit a large and growing chunk of America—supermarket chains realized they could pursue more lucrative customers with broader, deeper—and higher margin—arrays of packaged goods. By contrast, they saw Amazon as a growing threat. The online behemoth was a major force in the growing grocery home delivery movement. Amazon Mom, a diaper subscription program launched in 2010, offered 30 percent off diaper supplies, often with free two-day shipping.116 Supermarket News noted that the deal undercut just about everyone, including Walmart. “Thirty percent is a huge incentive to steal a trip and loyalty [from the supermarket] and try to build the mom’s [Amazon shopping basket] up,” noted an industry consultant.¹¹⁷

And it wasn’t just diapers, as Amazon began to experiment with home delivery of all sorts of groceries in the area surrounding its Seattle base. Called AmazonFresh, the program offered free delivery for orders of $75 or more (and $5.99 for orders under $75). Further, customers who lived in certain ZIP codes and who shopped once a week and scheduled deliveries on a specific day would receive free delivery with any purchase.¹¹⁸ The firm had experimented with grocery delivery since 2006, enlarging the service steadily over the subsequent five years. In early 2011 an article in Supermarket News reported that the company was mulling expansion of the service beyond the Seattle area. The article quoted an analyst who noted that Amazon could expand its grocery delivery “as far as it wants as long as it has enough of a distribution system, because there’s a limit on how far it can go from just one hub.”119 Amazon had, in fact, begun rapidly expanding its warehouses across the country.

Also in 2011 Amazon introduced its price-check app, which grocers realized was a major threat to their control over their store environment. An app that enabled customers to seek lower prices from Amazon as they moved through the aisles of their local supermarket could well result in increased grocery purchases outside the physical store. The widespread concern that permeated the retail department-store industry had now spread to the grocery industry. “Amazon is the biggest thing to hit retailing since Wal-Mart went national,” proclaimed the Journal of Commerce Online.¹²⁰ Investor’s Business Daily viewed physical retailing’s predicament quite ominously. “With e-commerce now a major share of core retail sales and growing rapidly, the writing is on the brick-and-mortar wall,” it said. “Adapt or die,” the periodical added.¹²¹

Many believed that the large chains Circuit City electronics and Borders books folded principally because an increasing number of shoppers used these physical stores just to “showroom”—they would go to a store to check prices before making their purchases online from Amazon or other e-tailers.¹²² The smartphone (used by about 42 percent of the U.S. population by 2011)¹²³ was a handy tool in helping consumers compare prices as they walked through the stores. Department-store executives recognized that they needed to make fundamental changes, reported Investor’s Business Daily. “To survive, traditional retailers such as Macy’s [and] Nordstrom . . . are adapting to a fast-changing landscape in which the old one-size-fits-all model no longer works with today’s cross channel shoppers.”124 An immediate response was to stop carrying products that shoppers could showroom. The trade press reported that stores began working with manufacturers to modify particular items slightly so that they would each be assigned a unique barcode. Department stores also reverted to their pre–World War II policy of carrying clothes with store labels. Macy’s, now the largest U.S. department-store operator, reported in 2012 that 43 percent of all the products in its aisles were either exclusive brands like Madonna Material Girl or in limited distribution and so hard to find elsewhere.¹²⁵ Many types of retailers, including supermarkets, could adopt this tactic, but it was not nearly enough; Amazon and other online sellers could come up with similar items that met or exceeded the quality of the physical retailers’ house brands, with quick tryout turnaround and at lower prices. Investor’s Business Daily reported that the house brand was just part of a more comprehensive strategy for competing with the Web and showrooming. Retailers, it noted, had “started to integrate the virtual and physical store worlds to give customers a similar experience.”¹²⁶

In the following chapters I will show that this integration is bringing us into a new era of shopping, with the retailing institution tilted firmly toward the ancient merchant impulse of discrimination. The difference is that, in the new world of the universal product code and the internet—and the ballooning amount of personal information that these technologies can generate on virtually any customer—it will be accomplished much more intensely. It will involve an interlocking system of new and old organizations committed to accumulating ever-increasing amounts of data on individual shoppers along with furthering the technologies to track and persuade them. All this information, and all these resources, are changing the industries’ understanding of the shoppers who move through their stores, of their stores’ layouts, and of the deals they can make with particular customers. The new world of the physical store is linked to a multifaceted approach to buying and selling, and it is challenging shoppers’ understanding of the marketplace, themselves, and even the traditional association between shopping and democratic social life that has become such a strong part of American culture.