Revolution Analytics, the R-based predictive analytics and data mining provider, now appears set to slide under the sprawling Microsoft tent. The announcement on January 23 gave few facts but did cause widespread speculation about the giant’s motivation. It needs R, goes the consensus, because it’s about to make a bigger push into BI this year or next.

The short answer to “why,” says veteran industry observer Dave Wells, “is that it gives them a quick way to get into the data science space without playing catch up through a long development cycle. But that is almost too obvious and easy an answer.”

For the deeper answer, we have to look into the Microsoft way of doing things. The company turns slowly, says Joe Caserta, president of New York-based Caserta Concepts, but when it does, the behemoth is “unstoppable.” Back when Oracle was adopting optimization for dimensional data models, for example, Microsoft was slow to respond. But when it finally did respond, “they completely leapfrogged Oracle.”

“They want to be the company you go to for predictive analytics,” Caserta says. To win, they need a strong analytics piece. IBM has SPSS, SAP has DataMiner, and SAS is a force unto itself. “Microsoft doesn’t have a play in that space.”

They have to join the club. The growth in the space threatens to leave them behind. “I’ve been in this business since 1996,” says Caserta, “and I’ve talked about R and predictive analytics in the past three years more than in the past 29 years combined. Microsoft would be crazy not to be in that game.”

Wayne Eckerson, principal consultant at Eckerson Group, says Revolution Analytics will help Microsoft beef up the analytics offerings of MSFT SQL Server Parallel Data Warehouse and HDInsight to compete with other database vendors which added analytics to their data processing platforms several years ago. “Microsoft is playing catch up,” says Eckerson.

For Revolution, says research analyst Mark Madsen, it’s a good exit. “It's hard to upsell from freely available R to a higher performance, easier-to-manage R. You're fighting the expectations that are set when a company starts with the open source ‘free as in beer’ R.”

But will R be enough to offer a good product? Madsen’s skeptical. Microsoft’s badly balkanized BI operation, spread across three groups with different goals and poor coordination, says Madsen, may trip over itself. “Each has its own view of the elephant.” Instead of BI tools, he says, the product is more like a BI toolkit. Building applications with it requires significant knowledge.