InformationWeek

When it's done right, creating a 360-degree view of the customer offers tangible competitive benefits. Companies that better understand their customers' preferences, past interactions, and demographic information can sell better--they know which customers are most important, most profitable, and most in need of service.

American Airlines Inc. keeps track of customers through its AAdvantage frequent-flier program, noting which customers fly most frequently and buy the most expensive fares. When a gold-or silver-level customer calls into the airline's frequent-flier center or checks in at a ticket counter, the customer-service representative knows who that customer is and how important he or she is, and can treat him or her accordingly. And if such a customer sends an E-mail to a new response system the airline is building, it will be one of the first answered.

Within a few months, American will include customer-profile information in its reservations systems so call-center reps and ticket-counter agents know their customers' most-recent experiences. "It'll build a closer relationship between the airline and the customer," says Elizabeth Crandall, American Airlines' managing director of personalized marketing. The system tracks basic information such as where customers fly and their preferred seating. American hopes to flesh that out with more important details--such as whether a passenger just had a bad flight or lost his or her luggage and needs to be treated with special care. The system will even suggest appropriate action, such as offering seating upgrades, bonus miles, or a free-flight voucher to a particularly harried traveler.

By understanding the sum of the customer's actions, says David Halek, consumer research director for marketing think tank Peppers and Rogers Group, a company can bring that customer into what his firm calls a learning relationship--one in which both parties gain. Companies find it easier to sell products, and customers feel that the business is attempting to meet their needs. Ultimately, it makes it more difficult for the customer to leave. "Over time, by having invested so much in the enterprise, the consumer is inclined to keep coming back," Halek says (see story, p. 52).

Analysis drawn from comprehensive data can go beyond personalization, cross-selling, or up-selling--it can transform a company's relationship with its customers. Chipshot.com Inc., a golf retail Web site in Sunnyvale, Calif., has been gradually building its customer-service offerings using Kana Communications Inc.'s Kana Response for E-mail management and Synchrony Communications Inc.'s eCRM product for call-center management. The next step will be to implement BroadVision Inc.'s E-commerce and personalization software so Chipshot.com can use customer data for tailored services. If the customer's Web-site visits center around titanium drivers, for example, the site can offer up content about big clubs, and on-site golf experts can provide advice suited to the user's long game.

But there are limits to analysis-driven personalization. For all the talk about one-to-one marketing, almost no one is building true one-to-one relationships with customers. For most businesses, it's simply not practical.

1-800-Flowers is piloting Prime Response Inc.'s PrimeVantage marketing software to analyze data and tailor campaigns aimed at particular customer segments. People who've bought flowers but not gifts from the company could be shown the other products the company sells. Registered users, who are among the company's most loyal customers, might be rewarded with discounts. The focus is on well-defined groups rather than individuals. "We might have 100 promotional campaigns, but not 8 million," says Paul Davis, the company's senior knowledge manager.

However, the groups are likely to get more specific, and more meaningful, given the kind of data the retailer can collect. A typical 1-800-Flowers customer leaves behind a bunch of personal information--his or her address and that of the recipient, payment information, marital status, the occasion for which the flowers are intended, even the words on a gift card. "We know if you're a heavy rose buyer, and you get in trouble a lot because you buy yellow ones," says 1-800-Flowers' Hage. But the company isn't acting on such detailed information yet, because Hage hasn't determined how.

For some companies, personalized attention is critical. Rittenhouse is analyzing the data accumulating in its Siebel system to help the company understand the differences among its customers, and act upon them--important because the brokerage firm's customers have widely varying portfolio needs. "There's a vast difference between a dot-com millionaire and somebody who recently inherited a sum of wealth," Crager says. "In the past, we would have treated them the same because they have the same net worth." But even though the two customers have the same account balance, the first is more likely to be a risk taker, and his or her portfolio should reflect that. Rittenhouse is looking at software from Epiphany Inc. that can catch such differences, then tailor portfolios and marketing campaigns appropriately.

That's the future. For now, Rittenhouse is attempting fine-grained customer analysis with predictive modeling software from @Risk Inc. The @Risk software, which Rittenhouse is testing in its consultant-adviser services group, can analyze customer behavior and predict which customers are liable to leave the company within three months if the company doesn't take action. @Risk evaluates small segments of customers who share similar characteristics, looking for clues in behavior such as a drop-off in calls or a transfer of holdings. If one customer in the segment leaves, the others are likely to leave as well. For instance, if someone calls several times a week, withdraws a portion of assets, and looks at performance data frequently, that tells Rittenhouse that person is a performance investor with a short-term focus who may be getting ready to bolt. So far, Crager says, the product has been 70% accurate.

But @Risk, which updates every night, doesn't determine only which customers are at risk; it also helps Rittenhouse decide what to do to keep them. The software feeds its data back into the Siebel application and suggests what action representatives should take. For example, in the case of a performance investor, Rittenhouse could point out the advantages of long-term investments, or build a high-performance portfolio that matches the customer's personality. If the customer's broker calls Rittenhouse, even to talk about another customer, Rittenhouse can warn the broker that the first customer may need attention.

Crager says Rittenhouse is using its customer-management systems to create an environment that tries to anticipate customer needs, before even the customer knows them. But although Rittenhouse has big ambitions for one-to-one marketing driven by analysis, Crager is the first to admit that his company is early in its efforts. Knowing what a 360-degree customer view should look like and actually being able to see the full picture are two different things.

In its search for the full picture, Harbor Freight Tools Inc. is gathering and analyzing data from a number of sources. The Camarillo, Calif., power-tool company has its roots in the catalog business and operates on multiple fronts--catalogs, retail stores, and the Web. It's gathering clickstream data from the 460,000 customers that visit its Web site each day, then analyzing it with software from DataSage Inc., now part of Vignette Corp. The tool company pulls transaction data from the point-of-sale systems in its 82 Harbor Freight stores throughout the West, and it culls information such as purchase history and tool preferences from catalog orders. After pooling all that data, the company analyzes it.

continued...page 3, 4
return to page 1