The Aisles Have Eyes

How Retailers Track Your Shopping, Strip Your Privacy, and Define Your Power

Joseph Turow

• Publisher: Yale University Press
• Year: 2017
• ISBN:
  9780300225075
• Topic: Business
• Subtopic: Consumer Behaviour
• Index: Business

The author of Media Today offers "a trenchant, timely, and troubling account of [retailers'] data-mining, in-store tracking, and predictive analytics" ( The Philadelphia Inquirer ). By one expert's prediction, within twenty years half of Americans will have body implants that tell retailers how they feel about specific products as they browse their local stores. The notion may be outlandish, but it reflects executives' drive to understand shoppers in the aisles with the same obsessive detail that they track us online. In fact, a hidden surveillance revolution is already taking place inside brick-and-mortar stores, where Americans still do most of their buying. Drawing on his interviews with retail executives, analysis of trade publications, and experiences at insider industry meetings, advertising and digital studies expert Joseph Turow pulls back the curtain on these trends, showing how a new hyper-competitive generation of merchants—including Macy's, Target, and Walmart—is already using data mining, in-store tracking, and predictive analytics to change the way we buy, undermine our privacy, and define our reputations.Eye-opening and timely, Turow's book is essential reading to understand the future of shopping. "Turow shows shopping today to be an exercise in unwitting self-revelation—and not only online."— The Wall Street Journal "Thoroughly researched and clearly presented with detailed evidence and fascinating peeks inside the retail industry. Much of this information is startling and even chilling, particularly when Turow shows how retail data-tracking can enable discrimination and societal stratification."— Publishers Weekly "Revealing... Valuable reading for shoppers and retailers alike."— Kirkus Reviews

1A FROG SLOWLY BOILED

It’s said that a frog placed in a boiling pot of water will escape, but if the water is slowly heated the frog will habituate and eventually die. Although scientists dispute the accuracy of this statement, no one in the audience of marketers objected—or even said it was ethically inappropriate—when a digital-advertising executive used the image in an off-the-cuff suggestion of how marketers ought to treat people in physical stores. The occasion was a conference called The Internet of Things: Shopping that the online trade journal MediaPost convened in Manhattan during August 2015. The speaker, vice president of one of the world’s largest interactive agencies, wasn’t invoking the frog image because he wanted to kill shoppers; he was addressing the concern that people would push back against beacon surveillance. Department stores, supermarkets, pharmacies, discount chains, and other retailers have among them already placed hundreds of thousands of these inexpensive devices throughout their stores.¹ If shoppers carry the right apps on their smartphones and have the correct technology turned on, the beacons will alert the merchants and they can send the shoppers personalized coupons or other messages associated with the goods in a beacon’s proximity. “We need to do a lot of hand-holding with our customers in the new environment,” he urged. The goal had to be to treat shoppers like “a frog in a pot of boiling water”: they had to be introduced to all the changes slowly so that they would come to consider them as a normal part of their lives.

Brandon Fischer, director of predictive insights at the influential GroupM Next consultancy, offered a view of the in-store surveillance future that seemed to embrace the frog comparison. In a keynote talk at the meeting, he predicted that by 2028 half of Americans (and by 2054 nearly all Americans) will carry in their bodies device implants that communicate with retailers as they walk down the aisles and inspect various items.² Based on how long you hold an item the retailer’s computers will tell whether or not you like it. Other signals from the implant will indicate whether you are nervous or cautious when you look at the price of the product you are holding. The analysis may lead, Fischer suggested, to a conclusion that a discount on the product may reduce your nervousness and lead you to purchase it. His argument was a blunt, optimistic case for biometric monitoring in stores. And just as with the frog-invoking executive, no one in the room protested. No one wondered, either, whether in just thirteen years people would realistically consider this activity natural.

The attendees’ nonchalance might seem strange, but the retailing business was changing so drastically and confusingly that such statements may well have seemed plausible. The message the marketers were hearing at the meeting and throughout their industry was that retailing is entering a new, hypercompetitive era with internet sellers. Brick-and-mortar merchants—the department stores, supermarkets, specialty stores, and chain stores that are still the center of the retailing universe—will succeed only if they figure out how to trace, quantify, profile, and discriminate among shoppers as never before. But for stores to survive this transition, shoppers will have to slowly learn to accept, even welcome, those eyes in the aisles as part of their natural environment—sort of like the frog in the pot of water.

This book is about how and why all of that is taking place, and the impact it is having on individuals as well as on society at large. I’ll show how a new generation of merchants—Walmart, Target, Macy’s, Stop & Shop, Safeway, Lord & Taylor, and many more—is working with a set of technology organizations to build a new future for physical retailing. They are reorganizing shopping to capture data about us through the very media we carry, even wear (such as a Fitbit). The goal is to routinely track us, store information about what we buy and when, and score us based on that and other information. Different forms of these activities are already under way in many stores. Retailers are using increasingly sophisticated electronic monitors, which show up first as experiments and then become ordinary elements of shopping. We typically feel their presence through personalized discounts often linked to our rewards cards. Depending on who you are and where you shop, you may already have experienced early versions of the following situations—scenarios some shoppers would consider wonderful:

• As you walk into an upscale department store, you may or may not realize that your phone signaled your arrival. The store cares because you belong to its loyalty program and have achieved high-value-customer status. Your presence is indicated to a store representative, whose tablet calls up your photo so she can recognize and greet you. The tablet also reveals which clothes you looked at on the store’s website during the past week as well as the clothes you clicked on when you accessed the store’s ads while visiting other websites during that period. Based on previous purchases and the information it has concerning your age, income, occupation, and family status, the store’s computer predicts which of those garments you will buy. It also suggests matching accessories, again based on your website visits, previous purchases, and the special predictive sauce that mixes these behaviors with demographic information. When you complete your shopping and go to pay, you are pleasantly surprised to find that the computer is rewarding your loyalty in the form of a 20 percent discount on your purchases that day.

• You enter your local supermarket, with the store’s app on your smartphone. The app instantly springs to life as you walk toward the first aisle. It retrieves your shopping history and loyalty score from the firm’s computers and links them to the shopping list and Web coupons you had loaded on the app at home. The computer analyzes this information and concludes that this is a stocking-up visit rather than a drop-in for only a few items. Complex personalization formulas, which evaluate your shopping list and your location in the store, present you with ideas about what to buy, recipes based on what’s in your cart, and discounts. The formulas also factor in information the supermarket has bought from data firms about your socioeconomic status, and assessments on where you are on various product buying cycles. Do your purchase patterns suggest you are at moderate risk for switching away from a particular shampoo? Does your increasingly less frequent purchase of a specific brand of diapers suggest a situation that can be countered with a $2 coupon (through a deal with the manufacturer), or is it likely your child is now potty-trained and so no longer needs them? Based on your shopping history, the formulas predict the extent of discounts that are needed to make you feel good about your shopping experience while also getting you to spend at least 10 percent more than you did the last time you came in to stock up.

• A similar scenario takes place in the big-box discount chain you visit often to buy household items. In addition to the information the supermarket used to send you messages and deals, this merchant has bought predictive data about your likes and dislikes based on the products you discuss on Facebook. The chain also bases its formulas for offering you discounts partly on an “influence” score it has bought from a company that evaluates the number of friends you have on social media and your degree of influence on them. For this trip you use the store’s app on your phone to scan the products you want to buy, incorporating the personalized discounts as you go. Your loyalty status and checkout history give you the ultimate reward: by using your phone to scan your purchases, you bypass the long checkout lines and instead simply push a button at a station near the store’s exit. There the retailer’s computers also compare your scanned items with your purchase history and conclude you have not stolen anything. No one searches your bags (as sometimes can happen at stores that allow product scanning throughout the store as opposed to a single designated checkout location near the exit), so you’re out superfast. No sweat.

Then again, depending on who you are and where you shop, you may already have experienced early versions of these personalized experiences many would consider unpleasant:

• No store representative greets you when you walk into the upscale department store, because your customer status doesn’t warrant it even though you belong to the store’s loyalty program. The store’s computer knows from your shopping history and background that you typically purchase clothes with greatly reduced sale prices, and that you are likely to continue doing this. Representatives therefore prefer not to spend time with you, but you don’t mind not being shepherded through the store; in fact, you rather like wandering alone. However, at checkout you see people in front of you joyfully surprised with 20 percent discounts, and you’re envious. If you were given an extra 20 percent off that already on-sale sweater you were admiring, you might be able to justify buying it.

• You live in a lower-income neighborhood and typically rely on a local independent grocer for all your shopping. This merchant doesn’t accept electronic coupons, and the paper ones you receive in the mail or encounter in the store don’t match your needs terribly well. You occasionally do go to the chain supermarket in a different part of town, and you’ve found that your smartphone app does give you some relevant offers. It seems, though, that the supermarket’s computer doesn’t know enough about you to give you the various good deals that you hear other customers discussing with friends as they circulate among the aisles. In fact, at the checkout you notice some shoppers getting great deals on products you would like to try, but you have to pay full price. Even a coupon for $2 off on one of those goods would be nice to have. You wonder if providing the app with more information about yourself would bring you better deals. But even if that would work, you don’t know how to do it.

• You stop at a convenience store to pick up a few things. Once inside you notice a sign stating that the store’s cameras use facial recognition technology to search databases for people who have criminal records. You actually have such a record, though it’s several years in the past. The presence of the data-retrieval software may or may not explain why an employee seems to be following you as you move through the store. It also may or may not explain the curt treatment from the clerk when you go to the register to pay, as well as his prolonged examination of the $20 bill you offer as payment.

Before you feel too relaxed because you are the recipient of great service and a lot of coupons, realize that the behind-the-scenes tracking may well have consequences you might not like. Retailers might hire statistical consultants to generate reports about your eating habits based on the food you buy or about your weight based on the clothes you look at online and in the store, or to develop more specific health prognostications based on the groceries and nonprescription drugs you purchase. Their portrait of you may result in some nice coupons for you to redeem now, but it may turn sour later as you age, as statistical formulas may well make unflattering inferences about you and your family. Consider, too, that some retailers sell or trade the information they compile about their customers; some even assign “attractiveness” scores to shoppers based on this data. The scores and the many points of information about you may affect the sorts of insurance offers, food and clothing advertisements, and travel deals you receive. And in the not-too-distant future the knowledge that companies have developed about you may lead news organizations to highlight, and even rewrite, certain stories for you, and advertisers to provide you access to certain pay television programs but not to others. Much of this will be happening—and so much is already happening—without your consent or knowledge.

Oddly, although these practices relate to the ongoing and widespread public discussion about privacy—many government hearings and papers, advocacy-organization reports, academic meetings, and popular press pieces discuss the ways online marketers and government agencies track citizens—retailers only barely figure in the debate. The shopping aisle has, in fact, received almost no attention even among academics who focus on the social implications of consumer surveillance—an unfortunate trend, because the traffic that retailers can track through those physical doors is huge. According to the Food Marketing Institute, in 2015 Americans made an average of 1.5 trips per week to a supermarket.³ The National Association of Convenience Stores (NACS) reports that in 2014 customers made nearly 160 million visits per day to the 152,794 convenience stores in the United States—58 billion visits per year.⁴ And according to the leading retail analysis company ShopperTrak, during November and December 2013 Americans paid 17.6 billion visits to malls, department stores, “big-box” stores such as Walmart and Target, and specialty retailers such as Express.⁵

Although visits to supermarkets and convenience stores have remained rather steady in recent years, foot traffic to the stores that ShopperTrak audits has decreased substantially; the 17.6 billion in 2013 had been 33 billion in 2010.⁶ Industry insiders generally believe that this drop is the result of increasing numbers of purchases over the internet. Because people can now shop electronically and have access to quick-delivery options, physical stores are competing with sellers not just in the same city or country, but from all over the United States, and even the world. What’s more, even when shopping in a physical store, customers access the internet from their smartphones to use as a competitive weapon: product ratings, price comparisons, comments of friends via social media, and ads from competitors all affect whether and how much people buy. For their part, most brick-and-mortar merchants have tied part of their fortunes to electronic sales, and in the process now can also successfully track and profile shoppers, largely without their permission. Yet despite the huge growth of online commerce in the past decade, numerous studies indicate that over 90 percent of product purchases in 2015 still occurred at checkout counters in physical stores, and few in the industry are close to suggesting that they will fade away in the near future.⁷ Industry experts do agree that brick-and-mortar merchants will succeed only by making the tracking abilities of their physical stores at least as good as or better than the virtual ones for targeting individuals with products and pricing. This includes using their data sources to create personalized messages for shoppers as they track them entering the store and proceeding through the aisles.

The implications are profound. The retail industry’s data-centered activities are restructuring the architecture of both physical and digital retailing as well as the relationship between the two in ways that turn enormous information gathering into something customers take for granted. To make shoppers they care about feel good about making purchases, merchants are fashioning new visions of “rewards” that remake the retail phrase “owning the customer” for the internet age. “This is an era of unprecedented change for retail,” Target’s chief financial officer told the New York Times in 2014. “In order to win,” he said, stores must keep “guests engaged with you as a business.”⁸ For Target and many other merchants, building relationships with individual shoppers today requires monitoring and discrimination. Retail monitoring involves gathering or purchasing information about shoppers’ backgrounds and activities with or without their explicit permission or knowledge. Retail discrimination has two meanings, numerical and prejudicial. In the first sense, discrimination is a human and organizational impulse to note differences among things and among people. In stores it means maintaining records on individuals and performing complex quantitative analyses on that data—for example, determining which incentives will get particular individuals to buy more goods. Prejudicial discrimination follows as a result of the value the retailer places on each individual based on perceived differences: the retailer will offer specific shoppers different discounts on particular products reflecting the data’s statistical portraits of those individuals and their families.

Part of retail discrimination is to identify customers deemed to have a relatively high “lifetime value” (a shopper’s lifetime being typically defined as five years). These profitable shoppers receive tailored deals aimed to keep them coming back. Customers on the lower-valued end of the shopping spectrum typically aren’t treated poorly; they may even get personalized discount offers in the hope that their value to the store might increase. But they will not enjoy anything like the attention and value the loyal customers enjoy. Moreover, some retailers downgrade the benefits of their loyalty program for customers judged to be of less value to the store based on the amounts they spend.⁹

These monitoring and discrimination activities usually fly under the public’s radar. An exception was a New York Times article discussing Target’s analysis of customer purchase patterns to identify pregnant customers. In the case of one family a father discovered his teenage daughter’s pregnancy only because the retailer had mailed her a package of pregnancy-related deals.¹⁰ These occasional press accounts no doubt cause at least some concern among those who also worry about contemporary government and corporate surveillance, but thi...