Riding the Marketing Information Wave

The phrase “getting close to the customer” now has a definite high-tech ring. Farsighted companies like American Airlines and R.J. Reynolds have gained a decisive competitive edge by building powerful customer information systems. Through such systems, these companies not only understand individual consumers better but also employ information to develop and market new products. But while the latest customer information systems support micromarketing and other alternatives to mass marketing, they are difficult and expensive to build. Because of the forbidding entry costs, some top managers question whether information technology can achieve true competitive miracles. Yet I believe that retailers and consumer goods companies can’t afford to ignore the new economies of scale that the best marketing information systems offer. Retailers, in particular, stand to gain the most from comprehensive customer information systems. Marketing IT will certainly accelerate the current shakeout in the retailing industry, much as the computerized reservation systems SABRE and Apollo revolutionized the airlines in the 1980s. Regardless of the obstacles, few marketers dispute the need to coordinate and integrate information. The growing number of market segments and the simultaneous increase in available products have made marketing much harder. Manufacturers are in a quandary about what to produce; retail merchandise buyers are overwhelmed by the task of product selection; and advertisers feel swamped trying to convey appropriate messages to so many market segments about so many products. Enter today’s customer information systems, which provide marketers with the means to cut through the confusion and to sort the most relevant data from the daily flood. Although many of the core technologies are still at a rudimentary stage, they are the foundation of an emerging generation of information systems. These sophisticated new systems can give companies a strategic advantage because they differ from traditional marketing in three ways: The scale of the database can be far greater than in the past, often including information on tens of millions of households. The depth of the information or the amount of information captured about each individual and household can also be much greater. Hundreds or even thousands of data items can be collected and analyzed on each household, including complete purchasing histories. The degree to which this information can be used as part of a highly automated business function bears little resemblance to the scattershot mass-marketing campaigns of the past. Indeed, companies with such customer information systems are now able to target small groups of highly responsive customers almost automatically. Micromarketing techniques like these are critical to the survival of large players, since they allow big companies to own niches in the manner of smaller, more flexible competitors. For example, it’s the efficient culling of so much information that allows The Great Atlantic & Pacific Tea Company (A&P), a major supermarket chain, both to target individual customers and to centralize buying for cheaper prices. And Fingerhut, the fourth largest U.S. mail-order company, has used its sophisticated customer information systems to succeed where traditional techniques regularly fail: selling merchandise on credit to households with less than $ 25,000 in annual income. Today’s information systems let big companies own niches the way small companies do. Then there are media companies like R.R. Donnelley & Sons, the world’s largest printer. As advertising dollars continue to shift to narrow-channel media like cable TV and specialized catalogs, Donnelley offers its clients everything from consumer and life-style data to customized individual publications. Such companies are grappling with the fact that mass advertising campaigns have become less and less effective in reaching diverse groups of consumers. Advertisers must now fight to establish the relevance of their products in an extremely noisy marketplace. Precisely because the market demands targeted promotions and advertising campaigns, companies need extensive customer information systems. And just as large companies benefited from past economies of scale in production, today’s large marketers are at an advantage. After all, the airlines, tobacco companies, and big retailers are the ones with the resources to invest in the right systems at the right time. But it’s not just a matter of size. Large companies without the vision to develop their own marketing IT—or the niche-oriented strategy of a successful small company—will drown in the wake of smarter competitors. Information Overload Effective marketing relies on a two-way information flow between the marketer and the prospect. Marketers must collect detailed demographic and life-style information about large numbers of consumers in order to determine effective market segments. Then they must integrate this mass of information into a concrete understanding of what products different consumers want and what they’re willing to pay. Effective marketing also means reaching the consumer with appropriate information in the form of advertising, store displays, and retail concepts. This includes positioning information to establish image, brand, and detailed product specifications. In particular, marketing new products requires an effective communication flow to inform the consumer about features and benefits. Yet when the number of market segments and the number of products dramatically increase, traditional means of exchanging this information become overloaded. Consider the retail food industry. The average grocery store now stocks around 20,000 items, with larger stores carrying two or three times that amount. Around 10,000 new items are introduced each year. The average supermarket now carries roughly twice the number of items it carried ten years ago. Given the enormous variety of weights, flavors, colors, and sizes, the total number of stock-keeping units at a supermarket may run to several hundred thousand. This creates an information crisis for retail managers and for manufacturers in the buying process. The supermarket industry’s approximately 550 corporate and regional buying offices are being attacked by an army of about 100,000 salespeople, manufacturers’ reps, and brokers selling some 60,000 different products. Moreover, the merchandise buyers deal with several hundred thousand price changes every year. One supermarket industry analyst refers to the 550 buying offices as the Khyber Pass of the food industry. It’s very difficult for this small number of people to make the best decisions about which products their customers want, how to price them, and which products to promote. Often these decisions are made not on the basis of what is best for the supermarket’s long-term relationship with its customers but on whether or not slotting allowances (payments for shelf space) or other promotional discounts can be wrested from manufacturers. One supermarket industry analyst refers to the 550 buying offices as the Khyber Pass of the food industry. Surveys of supermarket executives show that they expect technologies like point-of-sale scanning to help allocate shelf space and optimize product mix. Indeed, throughout retailing, the point-of-sale technologies introduced during the 1980s have helped stores cope with greater product assortment and variety while controlling inventory and avoiding shortages. But when you ponder the scale of the latest marketing information technology, it’s easy to see why companies are still wary of making the initial investment. Within a few years, customer information systems will handle tens of thousands of characters of information on tens of millions of customers, ultimately requiring trillions of characters of computer storage. Marketers can now get storage of up to hundreds of gigabytes of on-line data fairly inexpensively via massively parallel computer architectures provided by service companies like Teradata. Still, for many small retail or manufacturing companies, the cost of storage space alone makes new information systems hard to accept. Even more challenging is the corporate task of making this huge amount of data accessible and meaningful to the many marketing personnel who could benefit from it. Unlike previous marketing technologies (such as statistical analyses of demographic data or market segments), customer information systems can be used effectively by many different groups. With the right system, product marketers and brand managers can adjust prices and gain insight into new product possibilities, while advertising and marketing managers in the same company can develop promotions and produce personalized advertising. Recent commercial technologies like the workstations of client-server computing now enable different groups to use large customer information databases without becoming computer or statistics experts. Artificial intelligence systems, based on the emerging technology of neural networks, can sift through mountains of data and identify shopping trends and product opportunities, weigh promotional possibilities, segment groups of consumers, and present marketers with clear conclusions and suggestions. Once again, however, these systems have high start-up costs. The U.S. airline industry exemplifies the devastating leverage of customer information technology and the advantages of size in getting such expensive systems off the ground. Deregulation of the industry in the 1980s opened the door to more competitors and relaxed restrictions on routes and fares. As a result, the number of possible fares and routes skyrocketed. But because deregulation created information overload, customer reservation systems became the key to competitive survival. Today American Airlines’s SABRE contains 45 million fares and handles 40 million fare changes per month. Now the industry is dominated by just a few large airlines, the two largest being the pioneers of computerized reservation systems. SABRE and United’s Apollo, of course, cost hundreds of millions of dollars. Moreover, American and United had more than a decade of experience with their systems before deregulation hit. They rapidly changed fares and accurately targeted routes based on the information they had already collected, and their competitors simply couldn’t keep up. The analogue in consumer marketing to the proliferation of airline routes and fares is the flood of new products. Today’s torrent of products presents immediate challenges: finding customers for the new offerings and letting customers know efficiently that the products exist. All this puts a premium on gathering consumer information—not only to guide the designers, manufacturers, and marketers churning out these new offerings but also to shut out competitors in the future. Building Databases to Own the Customer Where marketers once aimed to build market share, they now aim to own a market niche. Small consumer goods manufacturers and many direct marketers have been particularly agile players in this regard, targeting niches in which larger companies can’t offer the same speed or service. And until recently, regional retailers were closer to the customers in their area, able to respond to local demand much more effectively than a nationwide chain could. But an increasing number of large retailers like The Great Atlantic & Pacific Tea Company are using technology, such as point-of-sale scanning and frequent-shopper programs, to own the customer. A&P has instituted a frequent-shopper program in several regions of the United States aimed at building a sophisticated customer database. Customers at participating A&P stores are invited to sign up for free Bonus Saver cards, which entitle them to check-cashing privileges. In addition, hundreds of items are marked with special discounts each week for holders of these cards. The discounts are obtained by showing the Bonus Saver card at the checkout counter, where a cashier runs it through a card-reading device connected to the cash register. At the same time that the customer is saving a few pennies, A&P is collecting valuable information. The card-reading device transmits the customer’s identity and a full list of his or her purchases to an in-store computer that calculates discounts and saves information on each customer’s purchases. This information is then transmitted weekly to a mainframe computer at A&P’s Montvale, New Jersey, headquarters. Building a customer database of such scale and depth will enable A&P to form long-term relationships with its customers and to focus its marketing efforts, gaining competitive advantage over smaller regional chains. (See the insert, “Invasion of Privacy: When Is Access to Customer Information Foul—or Fair?”) Invasion of Privacy: When Is Access to Customer Information Foul—or Fair? Now that automatic-dialing telemarketers have invaded most American households, the public has become increasingly aware of database marketing methods. Revelations of credit-bureau inaccuracies and the occasional misuse of credit-bureau information have prompted numerous consumer complaints. And in 1991, Lotus Development announced Lotus MarketPlace, a product that provided household information on CD-ROM, making widely available information that had routinely been used by large database marketers. As a result of this announcement, 30,000 people telephoned Lotus to complain about invasion of privacy. The company withdrew the product. With such public outcry over invasion of privacy, how viable are business strategies based on large customer databases? While the public’s concern is not to be taken lightly, it may also have unforeseen—and profitable—consequences. More stringent regulation of databases of personal information will, if anything, increase the strategic leverage these technologies offer. For one thing, such databases will become more expensive for companies without proprietary customer information systems. The key seems to be whether or not customers believe that they have voluntarily offered information to a company. For instance, if a customer buys a shirt at a department store, he or she has made a conscious choice to be in that store and, in some sense, to offer information through point-of-sale scanning or any other system that’s part of the transaction. Yet when Lotus proposed to sell a product to other companies with such customer information, people clearly felt they had lost control of how that information was used or exchanged. Because Lotus MarketPlace offered data in an accessible form, consumers thought that they would be subjected to even more harassing phone marketing or that private information would be revealed to other individuals. In fact, the direct marketing industry has established a number of self-policing initiatives to curb abuses, including the Mail Preference Service. This service allows consumers to opt out of receiving direct mail and to remove their names from mailing lists. Only a small percentage of households, however, participate in this program. Although many people hate the piles of junk mail that arrive in their mailboxes, it’s possible they still believe this use of customer information is relatively benign. After all, consumers still have a choice about whether or not they respond to direct-mail marketing. No one can force these prospects to open the envelopes or fill out questionnaires and sweepstakes forms. Still, many commentators believe that increased government oversight of personal information is on the way. The 1974 Privacy Act restricted government invasion of privacy, but it placed no such restrictions on business uses of personal information. Now bills have been introduced in the U.S. Congress to establish a Privacy Protection Commission that would determine the extent to which the privacy of individuals identified in customer information systems is protected. In state legislatures, over 300 bills have been introduced to regulate credit bureaus and over 1,000 bills to regulate telemarketing. A number of European countries, most notably Germany, have already enacted significant restrictions on the use of personal data. These restrictions apply not only to direct marketing issues but also to a wide range of routine business practices. In 1989, for example, the French government prevented Fiat from transferring French personnel files to Italy because Italian privacy standards weren’t as tough as those in France. In addition, the European Commission is currently formulating a Privacy Directive to be implemented in 1995 that will apply to personal data stored in computers in all European Community member countries. This initiative has been the object of intense lobbying by trade and consumer groups. In both the United States and Europe, personal information is likely to be regulated in the following three areas: Collection: The proposed EC directive requires that consumers be informed of all possible uses of personal data at the time that data is being collected. Use: Some proposals require that consumers be given the opportunity to opt out of direct marketing campaigns at any point. Other proposals state that any use of the data requires the consent of the individual. Even more stringent proposals and current German law aim to prohibit some uses of the data that are common practice today, such as exchanging customer lists in order to profile and target new customers. Transfer and sharing of data: Various proposals strive to eliminate the transfer of personal information from one company to another. The EC is also considering a version of information trade protection, which would prevent personal information from being transferred outside the EC. Yet while these regulations will eliminate some businesses and change the character of direct marketing, they will hardly stem the tide. Indeed, Germany has a robust direct marketing industry despite its restrictive legislation. Moreover, much of this regulation will strengthen the strategic advantage of the largest database marketers. First, restrictions on the collection and use of personal data will raise the cost of this data. Consumers will be offered discounts or other incentives to volunteer information. This will raise the level of investment necessary to build an effective marketing campaign, meaning that small and midsize competitors will find themselves at a disadvantage. Indeed, the information that Lotus couldn’t sell to small marketers is routinely used by large retailers and consumer goods manufacturers. Second, restrictions on the transfer and sharing of personal data will make information alliances more difficult to achieve. Smaller marketers won’t be able to rely on third-party data and analysis from credit bureaus, credit-card companies, and marketing database companies, as they have in the past. Theoretically, then, a large company like Sears, which sells a wide range of merchandise as well as credit-card, financial, and real-estate services, could achieve a distinct advantage. Given such a company’s access to different types of information, it could compile the “big picture” of an individual’s consumption behavior without resorting to third-party sources. Because A&P, unlike other top grocers, has recently embarked on a strategy of centralized buying, its in-depth customer information has become critical. In most other chains, purchasing and advertising decisions are made regionally or at the local store level. But by creating a centralized buying department to handle many product lines, A&P believes it can obtain greater clout with manufacturers, achieve greater efficiencies in inventory management, reduce the number of buying personnel, and sharpen advertising production. It is just these sorts of operational and logistical efficiencies that have enabled companies in other retail market segments—most notably Wal-Mart—to achieve such strategic leverage. The Bonus Saver program helps A&P avoid the risk that its centralized buying structure will make the chain insufficiently responsive to local market needs. The program allows central-office merchandise buyers to know the detailed characteristics of who is buying what products at which price in which store. By combining the frequent-shopper database with demographic and life-style data, central-office analysts can further understand customer needs. In cooperation with store managers, merchandise buyers at company headquarters select products and prices tailored to customer preferences at each store and region. An individual A&P store now reflects its locale, including shelves of ethnic foods, and, at the same time, it offers cheaper prices than a regional competitor. A&P now stocks each store according to the tastes and needs of local customers. In addition, A&P plans to use customer purchasing profiles to aim promotions at individual consumers. The Bonus Saver cards already allow A&P to offer individualized incentives based on purchasing histories: for example, the system may be programmed to offer a “loyalty discount” of 25 cents or more to regular shoppers, or Pepsi drinkers may be offered Coke coupons. Eventually, A&P may use advanced advertising technologies to reach households with direct marketing. Versioning management programs, for example, can rapidly assemble up to hundreds of versions of the same ad, switching and reassembling the products and prices over and over again. A number of consumer goods manufacturers have also begun direct marketing programs to build databases of extensive customer information through special promotional programs. Because cigarette makers R.J. Reynolds and Philip Morris face unique market conditions, they have developed some of the most advanced database marketing programs. The tobacco companies confront not only big retailers that try to force discounts but also a market of declining unit sales in the United States. Cigarette makers can’t advertise on television and are constantly threatened with losing their access to print media. Yet for ten years and at a cost of hundreds of millions of dollars, the tobacco companies have offered discount coupons, prizes, and free merchandise in exchange for consumer questionnaires. Using this information, cigarette marketers have built extensive databases on the demographics and preferences of most American smokers. This data helps marketers anticipate market-segment changes, test new products, and target promotions aimed at increasing brand loyalty or encouraging competitors’ customers to switch. For example, in 1987, Philip Morris conducted a blind taste test in which advertisements and flyers offered two free packs of an unnamed brand. Two million smokers responded by filling out a questionnaire. They each received two packs of Philip Morris’s Merit brand and follow-up mailings. As it turned out, the company found that half a million smokers had been converted to Merit cigarettes. The New Marketing Laboratory Building large databases and capturing in-depth customer information are not the only ways to reap true economies of scale in marketing information technology. Fingerhut, based in Minnetonka, Minnesota, has been successful largely because it has a scientifically managed marketing operation that is grounded in its customer information platform. Every catalog mailing and major promotional campaign at Fingerhut—there were nearly 150 in 1992—is based on statistically determined predictions about consumer behavior. Such predictive modeling is a good example of how customer information can be structured and sifted automatically to target promotions. Fingerhut’s customers would be rated as poor credit risks using traditional classification techniques. Yet Fingerhut, which serves 13 million customers, has grown up to 14 % annually, even during the recession. The past decade has seen rapid growth of the consumer merchandise and mail-order industry in the United States: up to $ 29 billion last year. Most of this growth, however, has been achieved by specialty catalogs like L.L. Bean. In the cut-throat market of nonspecialty merchandise sold to middle- and low-income consumers, only Fingerhut and a few others have prospered. Some, like J.C. Penney, have attempted to move upscale and specialize in fashion goods; others have forsaken the big-book catalogs for specialty books such as those that Montgomery Ward produces in a joint venture with Fingerhut; and others, such as Sears, have left the catalog business altogether. As for the size and depth of its database, Fingerhut captures as many as 1,400 pieces of information about a household. These include typical demographic items like income and home ownership, appliance ownership, and purchasing histories for various categories of products. Fingerhut’s statisticians have also determined reproducible relationships between individual characteristics and aspects of consumer behavior: who is most likely to respond to a given promotion, purchase given merchandise, or use given credit terms, and at what time of year. Consider this example of the Fingerhut approach to “magic moment” marketing. For every household, Fingerhut’s database stores the name, age, and sex of each child. The company produces a monthly series of birthday flyers, tailored to a specific age group and sex, which include selections of toys and gifts. Parents of children with upcoming birthdays receive an appropriately targeted flyer with a personalized message, such as, “Here’s wishing Sally a happy birthday in May. With your order, you’ll get special birthday gifts for her, PLUS three other free gifts.” This mail-order company thus accomplishes on a large scale what neighborhood toy shops have been doing with their best customers for years. In general, Fingerhut’s primary marketing programs are divided into three broad groups: front-end programs for acquiring new customers, transition programs for evaluating new purchasers and selecting those who may become valuable customers, and back-end programs for maximizing the value of proven long-term customers. This complex of programs serves as a kind of marketing laboratory. Fingerhut’s predictive models are used to design a series of tests that are performed on prospective and existing customers. Specifically designed marketing programs may test a customer’s creditworthiness or readiness to buy, or they may test different merchandise selections, promotional offers, or even envelope and catalog-wrap designs. In addition to selling merchandise, every marketing program provides additional information about individual customers, fine-tunes promotions and merchandise offerings, and continuously improves the predictive models themselves. Fingerhut’s front-end marketing programs include ads in newspapers and magazines, television commercials, postcard mailings, promotional mailings, and specialized catalogs. The company’s predictive models play a central role in determining who receives front-end promotions. Most of these initial mailings are sent to names obtained from lists sold by other companies, including magazines, financial companies, and other retailers. Customers who respond to these front-end promotions then become prospects for Fingerhut’s transition programs, which sort out who will be responsive long-term customers and good credit risks. Assessment of creditworthiness is critical to Fingerhut’s profitability and market position. Even selling heavily to households with less than $ 25,000 in income, Fingerhut offers extensive credit terms on almost all merchandise to its best customers. Few competitors come close to Fingerhut in offering credit, and this capability provides unique market leverage. The transition programs are essential to sorting out the good risks and valuable long-term prospects. Customers are mailed a series of offerings with progressively more liberal credit terms. Risk is minimized by controlling the terms offered and the value of merchandise until the customer proves to be reliable. And once customers qualify as good risks, they become a prospect for Fingerhut’s back-end marketing. Of the company’s annual revenue, 80 % comes from existing customers; in 1992, sales to existing customers grew by 12 % despite the recession. The back-end programs are designed to maximize the lifetime value obtained from these customers. This is done by offering the right merchandise to the right customers at the right time. Customers are segmented into different groups based on their purchasing histories and other individual information. Different groups are then targeted for promotions and specialty catalogs. In 1992, Fingerhut mailed around 75 specialty catalogs, including outdoor living, electronics, fine jewelry, domestics and housewares, apparel, unique gifts, active wear, juvenile apparel and toys, and products for senior citizens. The catalogs are further customized with bound-in promotional offers and personalized messages printed on the wraps, which reflect in-depth information about individual customers. One such message might thank Mrs. Smith for her recent purchase of a VCR and recognize her fourteenth anniversary as a Fingerhut customer. A Fingerhut catalog will thank Mrs. Smith for her VCR purchase and recognize her fourteenth year as a customer. With its reliance on technology, specifically in the development of predictive models, Fingerhut’s marketing contrasts sharply with that of other retailers, consumer goods manufacturers, and even many mail-order companies. Yet Fingerhut hasn’t achieved all the advantages to be reaped from sophisticated customer information systems. For example, the company has just now established the electronic links with merchandise suppliers that can provide a basis for entirely new ways of selecting merchandise, streamlining inventory, and developing new products. The Curse of Irrelevant Advertising The efforts of Fingerhut, A&P, and others to forge closer ties with customers through marketing technology come at a time when many marketers are frustrated by saturated advertising channels. Advertising expenditures more than doubled during the 1980s; a typical consumer is now barraged with thousands of advertising messages every day. Yet advertising hasn’t kept pace with the number of new products. The problem is not too much advertising but that too little of it is relevant to any particular consumer. As an old marketing joke goes, junk mail is just poorly targeted direct mail. Since advertising noise levels have made it increasingly difficult to differentiate brands—and retail merchandise buyers have become overwhelmed with the selection—manufacturers have shifted their marketing budgets drastically from advertising to promotional discounts and slotting allowances, which in effect buy shelf space for their products. Yet this emphasis on price promotion only further undermines brand loyalty and makes it more difficult for a manufacturer to establish image and position. To break through the barrage of contemporary advertising, smart manufacturers are using customer information technologies to target promotional and marketing activity to ever smaller market segments—another form of owning the customer. Advertisers are looking for new ways to reach individuals more effectively through “narrowcasting” rather than traditional broadcast media like TV commercials and saturation newspaper ads. Cable TV advertising, in-store coupons, and special monitors that display advertisements in doctor’s offices, cash-register waiting lines, and airport lounges all exemplify the new narrowcasting possibilities. The introduction of interactive kiosks that let shoppers browse through video presentations of products, the imminent availability of 500 cable channels, and the growth of interactive home TV shopping go a step further into the future—a future in which the structure of today’s advertising and media industries is bound to be substantially different. One media company that has used leading-edge technology to take strategic advantage of the narrowcasting trend is the printer R.R. Donnelley & Sons. Despite the recession and large postal-rate increases, Donnelley has continued to rack up better than 10 % annual sales increases, including strong growth in its largest segments: catalogs, advertising inserts, and magazine printing. Donnelley offers its customers a wide range of services that go far beyond the traditional business of putting ink on paper. This printer will manage a customer’s database. Alternatively, it will supply its own consumer and life-style data to customers. It provides services to segment and target customer’s mailing lists. And Donnelley offers specialized printing facilities that make possible highly customized individual publications. Donnelley’s Selectronic printing and binding technology allows a catalog marketer to offer different merchandise to different households. Or a magazine publisher can offer different editorial and advertising content to different subscribers. In the past, direct marketing was prohibitively expensive for small groups. But new technologies like Donnelley’s Selectronic Services help control the costs. One of the first customers to take advantage of the Selectronic process was the Farm Journal. This publication began collecting in-depth reader information over two decades ago, including demographics, farm size, type of ownership, type of crops, and type and quantity of livestock. Such data allowed the magazine to provide tailored editorial and advertising content to small groups of subscribers. Donnelley now prints as many as 8,000 different editions of the Farm Journal per month. R.R. Donnelley now prints up to 8,000 different editions of the Farm Journal every month. Such examples merely indicate the shape of things to come. In part, that’s because many required technologies—like versioning systems and new printing press technologies that allow for two-color personalized messages and higher quality custom printing—are just emerging. But as new technologies become practical and as more customer information is gathered over the next decade, advanced marketers will proceed much further in building databases of great scale and depth. There will also be advances in the degree to which targeted promotions and advertising become automated. Consider what’s likely to evolve: with a wealth of fresh consumer data analyzed in powerful new ways, retailers and manufacturers will be able to define their target customers not just by demographic characteristics but also by complex clusters of actual purchasing behavior. They’ll be able to probe almost continuously and test for new product opportunities. Promotions, discounts, and advertising will be highly tailored to different consumers. The customer will see a level of service that’s impossible today, one that offers the broadest selection of specialized products yet also makes shopping a less burdensome task by serving the needs of the individual. A New Balance of Power The marketing future will undoubtedly look different in another respect as well: customer information technologies will change the relative roles of retailers, manufacturers, and media companies. Retailers have a built-in advantage because they can directly measure customer response and get first crack at the broadest range of information. Indeed, point-of-sale scanning systems have already played a significant role in shifting power from manufacturers to retailers. Some manufacturers, such as the tobacco companies, will be able to build substantial customer information systems that may help them to create customer demand for their products. But overall, the balance of power has tilted toward retailers, as large innovators like Wal-Mart have amply proven. Through centralized buying that allows for cheaper consumer prices, along with the high-quality service of a local store, Wal-Mart and others have turned up the pressure on manufacturers. Many retailers will also target new product possibilities with lines of private-label goods, further reducing the power of brand manufacturers. Manufacturers can also expect continued pressure from retailers to provide promotions and discounts, although the nature of such deals will change. Some manufacturers like Procter & Gamble have been replacing promotional discounts with “everyday low prices.” But extensive customer information systems will dramatically change the use of promotional discounts yet again, reducing across-the-board promotions and providing steeper but highly targeted discounts. With emerging customer information technologies, it’s possible for both manufacturers and retailers to offer discounts only to those customers for whom it pays—that is, those who are cost sensitive enough to switch brands or shopping habits permanently. Most important, the balance of power between large and small companies will change. As customer information technology becomes more prevalent, only those companies that can invest the resources and show technological leadership will succeed. As in the airline business, proprietary information and technologies may be used to erect barriers to competition. Complex customer information systems may require investments of up to $ 100 million, and it may take years to build the technology and collect the data. Unless they are farsighted, daring, or well-heeled, small or even mid-size competitors will find it difficult, if not impossible, to play this game. Customer information systems will provoke further consolidation in retailing, with larger, more technically sophisticated chains continuing to gain market share. But many large retailers will come under pressure as well. Top retail managers understand the benefit of $ 100 million investments in bricks and mortar but all too often don’t see the need for similar investments in technology. The long-term legacy of Robert Campeau and other real-estate-leveraged investors of the 1980s may not be that they drove viable retailers into bankruptcy; rather, some surviving companies have been left without the resources to make the critical technology investments needed in the 1990s. For consumer goods manufacturers, of course, customer information systems represent a basic shift in the source of competitive advantage. In the era of mass production and mass marketing, competitive advantage sprang from economies of scale in production. But during the 1980s, with the increased demand for new products, this economic logic was reversed. Flexible manufacturing techniques allowed efficient production of short runs—and short, specialized runs were what the market demanded. Consequently, the average size of manufacturing organizations sharply declined. Today many pundits believe the future belongs to small business. Yet the use of expanded customer information systems will reverse the logic once again. In fact, the current changes in marketing don’t apply to the information systems that support “small,” targeted strategies. Now the pendulum is swinging back toward the large companies that will be able to afford the most advanced customer information systems, provided these companies are visionary enough to adopt them. Big companies like A&P and Procter & Gamble will derive new economies of scale from their marketing information technology, not their production facilities, retail space, or mass-marketing campaigns. It’s true that small companies without such extensive resources can still rely on less sophisticated technologies, as long as their businesses occupy a clearly defined niche. But for small and midsize companies that aren’t following a niche strategy, the rapid advances in marketing IT pose the biggest threat. The survivors in today’s shifting marketplace will be the companies that react fast and target the right customers. In addition, while there’s a strong push for information alliances between manufacturers, retailers, credit-card companies, and other parties that collect customer data, exclusive alliances don’t exist yet. If they ever do, such alliances may give small retailers and manufacturers the opportunity to reap the benefits of customer information systems by teaming up with larger partners. Yet acquiring information for marketing purposes is likely to become increasingly expensive, especially given the public’s growing concern over invasion of privacy. Still, smaller players can hook up with sophisticated service and media companies like R.R. Donnelley. Even Fingerhut pays for customer information from other companies to compile lists of new prospects. And travel agents have benefited enormously from SABRE and Apollo, not faded into obsolescence as predicted. In a jumbled information economy of niche players and electronic shoppers, only one thing is certain: the current revolution in customer information technology is fundamentally changing the nature of competition.

Riding the Marketing Information Wave

Partner Center