It's no secret that e-commerce has turned retailing on its head. The phenomenal and oft-mentioned success of Amazon.com has forced traditional retailers into the sometimes uncomfortable position of tackling a medium both unfamiliar and money-absorbing. Companies like Amazon.com that sell their wares exclusively in cyberspace are faced with the obvious challenges of building systems from the ground up that can scale to meet customer demand in real time. But for organizations with established catalog or brick-and-mortar sales channels, or both, figuring out how to integrate e-business with an existing business and customer base presents more complex hurdles.
What has emerged as perhaps the crucial element of meeting e-commerce head on is the creation of a so-called unified view of the customer. In other words, a customer's on-line, catalog, and store-based retail activities are tracked, stored, and consolidated in one place, which results in a better understanding of buying patterns and a higher level of customer service. Companies employ business intelligence solutions to help build this customer knowledge base. Combined with such approaches as interactive relationship marketing, some companies are already building potent marketing and operational capabilities.
"Things like business intelligence and customer relationship management gets companies to the depth of information and analysis they need to understand their customers," says Erin Kinikin, an analyst with Giga Information Group, based in Cambridge, Mass. "E-commerce lets them use that information to provide rapid interaction with and feedback from customers, which is something that just wasn't possible before."
One company that has taken on the task of building a top-notch e-commerce presence using a business-intelligence system is Freeport, Maine-based L.L. Bean. The privately held 87-year-old company, which posted sales of more than $1 billion in 1999, is perhaps best known for its mail-order catalog business and outdoorsy product line that includes clothing, sports gear, and home accessories.
Top management at L.L. Bean realized early on that an e-commerce channel would be a crucial part of their strategy moving forward. The result of that decision was llbean.com, which was launched more than two years ago and through which customers can choose from nearly 1,000 of the 16,000 products L.L. Bean offers in its print catalog. Since its inception, the online portion of the business has been growing at a rate of 150 to 200 percent a year, according to Doug Faherty, L.L. Bean's director of customer initiatives.
As important as it was to build a Web-based store, L.L. Bean wanted the new sales channel to augment rather than cannibalize existing business. This meant ramping up its business-intelligence system in order to understand the precise impact of the Web site on its sales and customer base. To that end, online analytical processing (OLAP) software from Ottawa, Canada-based Cognos and analytical software from Cary, N.C.-based SAS Institute, along with some home-grown programs, was put in place. In addition, SP parallel processors from IBM were used to create a "super" computer that could provide better performance. "We were looking to have data from all three sources--catalog, retail, and e-commerce--in one place," says Faherty.
One of the first tasks Faherty and his team faced in implementing the new system was moving data off the company's mainframe and on to data warehouses that sit on the IBM-enhanced server hardware. The idea was that moving data off the mainframe would make data access faster and easier, and at the same time improve data quality, ultimately leading to better reports.
All customer and order information is still sent to the mainframe initially. The customer and purchasing information is forwarded to the data warehouse where it can be merged with existing demographic and historical data. Currently, the company keeps track of such things as how many times customers have been contacted, from which retail source their orders originate, the number of promotions (both electronically and by direct mail) received, and, of course, the items ordered.
While getting the information into a usable format is important, the ability to run the most useful reports against that information is critical. The company is in the process of moving from legacy reports written in Cobol to using the reporting tools from Cognos, which Faherty says are easier to create and manipulate.
Despite these obstacles, Faherty says, L.L. Bean has made significant process in terms of measuring how well its Web initiatives have fared compared with its long-standing catalog business. It knows which of its customers are comfortable shopping on the Web versus other channels. In addition, the new systems have given the company tremendous insight into the synergies between its e-commerce site and catalog business as well as the ways in which they cannibalize each other. "In this day and age, we have got to understand how e-commerce works with and against our traditional business," says Faherty.
Far from alone
L.L. Bean is far from alone when it comes to figuring out the e-business equation. According to Ben Barnes, general manager of IBM's business intelligence group, L.L. Bean's experience provides a crystalline example of what established retail organizations must do to pull their business into the electronic age. First, as L.L. Bean has done, companies must build a consolidated business intelligence warehouse that provides a single view of the customer, a notion the Giga Information Group's Kinikin agrees with wholeheartedly. "Companies have spent the 1990s getting to be more efficient and now they have to be smarter," says Kinikin. "A unified view of customers is a baseline requirement for being smarter."
While the so-called dot.com organizations or e-tailers may have a slight advantage in the short term, at least from an information technology perspective, Barnes and Kinikin believe that larger organization can more than compete thanks to their established brands, customer bases, and reputations. They must face up to the initial challenge required to put business intelligence to work for them. "Building a consolidated customer view is the heavy lifting, it's definitely tough," notes Barnes. "But once you get that done, it's so powerful; you can begin to really understand your customers and the types of services that they want."