Data Mining: New Economy Wizardry?

Revitalized as CRM's competitive edge, "customer analytics" are hot

Is it safe to say that "irrational exuberance" has become merely rational? While the B2B arena still floats a little higher than reality, many dot-com valuations look like downright bargains. In dot-com circles, profitability is now uneasily referred to as "the p-word." But weaned on unlimited capital, do dot-coms know how to get there? Data mining is the elixir for which many are reaching. And with good reason: That is, as long as the choice is a rational decision.

Data mining -- really a merry band of technologies that includes artificial intelligence (AI), predictive modeling, pattern recognition, and visualization -- is a core topic of this magazine. However, the bubble called the data mining "industry" burst (or perhaps gently deflated) a year or two ago. Silver bullet misfires and algorithm suites from 800-pound gorillas such as SAS Institute and IBM caused most of the more than 100 tools that once existed to be acquired, discontinued, or subsumed into other solutions.

However, this evolutionary process did not spell the end: far from it. If anything, data mining is becoming embedded ever deeper into cornerstone applications of the intelligent enterprise, such as customer relationship management (CRM) and supply/demand planning. Software products from such companies as i2 Technologies and Selectica, while perhaps not precisely data mining, nonetheless heavily use predictive modeling based on AI. Advanced enterprise application integration (EAI) products have their hands full supporting new technologies such as Java and XML, but will likely turn to data mining to sharpen performance management and enable federated data warehouses.

Nowhere is the opportunity greater for data mining software and services than in the red-hot CRM industry. While Siebel Systems, Oracle, and Broadvision duke it out over which company can field the market leading e-business/CRM suite, other contestants are definitely worth watching. These include Unica Corp., Xchange Applications, E.piphany, and thinkAnalytics. Most either entered the field with strong "customer analytics" (that is, data mining) capabilities or recently acquired such software. The high-end technology leader could turn out to be Blue Martini, which offers a suite that fully integrates e-business with serious customer analytics.

None of this stuff is inexpensive; users must beware being sold the same silver bullet, but in a different guise. With a few weeks to soak in knowledge from attending KDD 2000 in Boston, I hope to have more to say next time. While a little exuberant, data miners at the conference were quick to point out that this stuff does not work miracles overnight. Yet, as dot-coms breathlessly reach for mass customization, customer retention, and increasing share of wallet, no wonder they see customer analytics as the stuff profitability dreams are made of.