Total Pageviews

Tuesday, April 2, 2013

Why Big Data Is I.B.M.’s Friend

For big sellers of computer technology to business that are looking to stay relevant, there’s no friend quite like Big Data.

Things like cloud computing and mobile technologies are changes in the ways that computing is done. That makes them a threat to an incumbent’s legacy businesses selling, say, computer servers, laptops, or databases. Big Data, however, is a trend that relies largely on the digital information already in a company’s system. The company that got it there in the first place stands a good chance of helping a company make sense of it.

Case in point: I.B.M. just announced a series of technologies intended to make data analysis faster and more powerful. That is probably good news to Big Blue’s core customer base of large corporations, which are increasingly drowning in digital information from both the Web and the physical world. A good portion of that data is already derived from and running on I.B.M. technology.

I.B.M.’s offerings include an easier way to load corporate data into an I.B.M. machine and produce analysis quickly; an analysis “accelerator,” which can speed the production of analytics reports, I.B.M. says; and a Big Data “appliance,” or combination of hardware and software, that enables companies to posit questions, then have them answered automatically when the appropriate data comes along.

“Queries that took two days to execute can be done in three minutes” using the accelerator, said Bob Picciano, I.B.M.’s general manager of information management. The appliance, he added, “lets you ‘persist’ a question. That means you can know what to ask, without knowing about when you can ask it.”

A customer cited by I.B.M., BNSF Railway, was said to have had a 100-fold improvement and a 90 percent drop in data storage consumption using the new technologies. Mr. Picciano suggested that the analysis technology would be useful for an entity like a national retailer “looking for information patterns around occasions of sales over years of data.”

Mr. Picciano said that, after looking at “thousands of customer engagements,” his company has found five major uses for Big Data. In order of frequency, they are: retail, which wants better customer understanding; security threats and fraud detection; gaining efficiency in the operation of information technology; analyzing information from other networked machines, like sensors on electricity grids; and incorporating new data sources, like Twitter, into existing databases.

That is a fair portrait of needs among I.B.M. customers. Big Data, as sold by I.B.M., supposedly helps the large companies figure out how to understand customers that the big shops see only at a distance from headquarters. It can also be good at helping customers better run the expensive equipment they bought from I.B.M., among others.

In all, it is the kind of impressive technology that appeals particularly to companies with years of data on their hands and problems that stretch across nationwide chains of stores, or networks of sensors and computing systems.