One of the cruelest myths of the Knowledge Age is that it is easy to excavate business value from the vast deposits of corporate data. In reality, the integration and analysis of data stored in various information systems can be expensive and time-consuming. Digital Archeology's Discovery Suite 2.0 offers a solution that may ease the chore.
Data warehousing was supposed to have eliminated this problem, particularly in the case of sales, accounting and manufacturing information, which may be the most time-critical data a company needs. It doesn't always work that way. As data-retrieval capabilities have evolved, new repositories often are incompatible with the old. Also in this era of mergers and acquisitions, large companies newly cobbled together from smaller entities frequently confront conflicting or incompatible systems.
Increasingly, companies looking for fast access to strategic data have turned to data mining, using software designed to access multiple databases simultaneously. Yet this technology has not fulfilled its promise, according to Alexander Linden, senior lead analyst for data mining at the Gartner Group in Frankfurt, Germany.
"Data mining is statistical modeling--interesting but hardly revolutionary," Linden says. He acknowledges, however, that data mining can produce valuable knowledge when properly executed. "Buy a book from Amazon.com, and the company will present you with four other titles purchased by people with tastes similar to yours. Amazon earns an additional 5 to 10 percent in revenue because of this cross-selling feature."
Digital Archaeology Corp. of Lenexa, Kansas, proposes to make data mining more effective with back-end tools that handle disparate data better and front-end tools that make it easier to query the data.
For example, iTravel Smart, an online corporate travel broker, was concerned that visitors to its Web site were not buying tickets. Data mining showed they actually were, but the purchases were being made via the company's toll-free number or from affiliated travel agents because of worries over credit-card security on the Web. "Data mining revealed the pattern of consumer purchases and the source of their concern in a matter of hours," says David Frankland, Digital Archaeology CEO.
Digital Archaeology's Discovery Suite 2.0 is a group of customer-centric applications that analyze electronic transaction information, online demographics and clickstream data as well as traditional customer and sales information. Its purpose is to personalize product marketing according to customer preferences and buying habits. Through its interface, users form queries based on a supplied library of equation objects.
When William Venable, director of data-based marketing strategy for UMB Bank Financial Corp in Kansas City, Mo., used Discovery Suite, it took just 15 minutes to analyze 750,000 transactions--not on the bank's Web site but at its automated teller machines. "My goal was to convert noncustomers using UMB ATMs to customers by offering them low-interest home loans and other reduced-fee services," Venable recalls. "But I did not want to cannibalize our existing business."
Discovery Suite allowed him to compare data across disparate data sets so the bank could identify the specific needs of individuals. Today, UMB banks put display screens next to their ATMs so when a customer inserts his or her bank card, the screen will begin touting programs the bank thinks the customer is likely to purchase.
Digital Archaeology's primary focus is analysis of online customer behavior. Its platform integrates clickstream data with information from other sources--registration databases, call center information, transaction data from offline sales and third-party demographic information, for example--and distributes the analysis throughout the organization by means of Web-based decision support applications. Though some of these functions overlap with other data mining tools, Discovery Suite can be used to prepare data prior to feeding into other tools, claims Doug Dickerson, senior product manager for Digital Archaeology.
X-Set, the technology underlying Discovery Suite, enables the application to access multiple data sets directly instead of relying on the structure-centric approach used by relational databases, according to Dickerson. "Instead of the database structure determining the rules for questions that can be asked of the data, with X-Set the questions that are asked of the data determines the underlying structure," he says.