"What if you could quantify the world's consumer behavior," asks a company video, "and use it to make different and better decisions than you ever have before?" This is the question that has been asked over the past decade by a company called DemandTec, a manufacturer of retail analytics software for analyzing economic and behavioral trends around consumer purchasing decisions.

The idea is to change the conventional way manufacturers engineer their product lifecycles: first by developing it, then by marketing it upon completion. Those cycles have historically been separate; DemandTec's idea is to merge them, to make businesses make strategic merchandising decisions (e.g., "How about marking that tablet down to a hundred bucks?") rather than marketing corrections (e.g., "What if we changed that tablet's promotional slogan?"). It sounds like an idea IBM could wrap its head around. Today, that's exactly what IBM did.

Through its own blend of organic development with mergers and acquisitions - the most recent being trade investment management software maker M-Factor just last March - DemandTec is said by analysts to be among the biggest brands in its product category. IBM is paying $440 million for DemandTec in an all-cash deal that's typical of IBM's acquisition pattern (back in 1996 it paid $743 million for Tivoli Systems).

Today's deal indicates that IBM's on-again, off-again affair with business analysis services is on again. Back in 1996, the company began testing the waters of Web analytics with a product called SurfAid. Then five years ago, deciding it would cost too much to continue competing directly with bona fide consulting services, IBM sold SurfAid to a company called CoreMetrics. That company took SurfAid like a fumbled football, and scored big. Meanwhile, in 2009 Adobe acquired Omniture, in a $1.8 billion deal that a certain individual thought might be overpaying.

In August 2010, IBM reversed course and acquired Coremetrics, in so doing bringing SurfAid back into what the company calls its Smarter Commerce portfolio. DemandTec will find its new home within that same portfolio, but its purview is not Internet usage behavior. It's purchasing behavior, both online and in traditional retail markets.

"DemandTec's platform for customer-centric decision making, combined with advanced customer analytics and marketing expertise from our partners, enables retailers to gain new insights into customer loyalty," reads a recent DemandTec product brochure. "For example, retailers can analyze market baskets and product affinities to understand the categories, brands, and items that are associated with the purchasing behavior of the most loyal and most profitable customer segments. They can also perform detailed analysis of customer loyalty data to identify behavior that is predictive of customer attrition, and then apply prediction factors to existing customers to identify customers that are at risk of defecting."

The term "market baskets" in this instance is not a euphemism; it means the things that a consumer purchases together in one batch. DemandTec's data doesn't come from extensive polling of customer satisfaction, but rather by a measurement of what products get purchased as a result of certain promotions, which products get purchased along with others, how effective coupons are at driving sales, whether a consumer experiments next time with a different brand in the same product category, and data of that nature. And using consumer loyalty data, DemandTec's SaaS-based software can point the way to the most effective promotion a manufacturer or vendor may make, for a given product line, at a select subset of outlets, in a particular geographic area.

Sounds like IBM territory to me. And in light of the $440 million price tag, I'm wondering whether IBM used DemandTec's own software in projecting the best price.