RealTime IT News

IBM Talks Up 'Smart Analytics' Strategy

At the same time that it's announcing the $1.2 billion acquisition of analytics player SPSS, IBM today unveiled its plans for an integrated analytics offering called the IBM Smart Analytics System.

The efforts build on IBM's (NYSE: IBM) unveiling earlier this year of its Business Analytics and Optimization (BAO) massively parallel technology designed to be able to handle huge data sets.

"Companies are constantly challenged by the volume of data, variety of information and velocity required for decision-making. Leveraging analytics capabilities can help turn this increasing mountain of data into predictive intelligence to help both business managers and IT analysts run new digital infrastructures smarter, faster and more efficiently," Ambuj Goyal, IBM general manager for information management, said in a statement.

"The announcement marries software and hardware plus IBM's world-class consultants and delivers real results in days, not months or years," Jeff Jonas, IBM chief scientist for analytics, told InternetNews.com.

While Jonas said the system was optimized for an IBM-only system, known as a "blue stack," he said that the offering is based on IBM DB2, "so any product that can talk to SQL can get in and get the data out."

"It's difficult for customers to buy five or eight different products from different companies because it's tough to make them work together," he added. "The expertise required is non-trivial."

Jonas said IBM isn't just offering hardware plus software. "This system brings an asset to bear: 4,000 consultants dedicated to our BAO offering. They can walk in the door with something that's off the shelf and pre-optimized. Our professional services division is deploying systems and tailoring them to industries," he said.

Big blue data sets

So far, the plan is receiving the thumbs-up from industry insiders.

"Will it help lower the costs or lower some of the complexity? I think the answer the answer is absolutely," Claudia Imhoff, president and founder of business intelligence consultancy Intelligent Solutions, told InternetNews.com. "Anything you can do to pre-configure, pre-optimize, pre-build ... is going to make it simpler to install and maintain and so forth."

"They took all of the system sizing and querying of the componentry and installation... and all that stuff and kind of done it for you and put it into their smart analysis system. A lot of the hard, upfront work has already been done. That is certainly going to result in a faster time of deployment. "

The product supports tens and even hundreds of terabytes of data, Bernie Spang, director of product marketing at IBM, told InternetNews.com. It can analyze those massive data sets, too.

That's likely to be an important selling point, with enterprise systems increasingly challenged by the sheer volume of historical data they must maintain. While they may add terabytes each day, it is the trends analysis rather than real-time tracking that tasks servers and systems.

"All industries need both real time and historical data," Spang said. "The interdependency of both can be seen in retail in the need to understand long term trends as well as the immediate impact of interactions with customers."

One customer includes the Blue Cross Blue Shield Association, according to Spang, who said the group's plan includes "assembling the largest health intelligence database representing their 54 million members nationwide."

Spang said Blue Cross Blue Shield will also analyze trends for fraud and crime prevention. The offering can also help patients thanks to real-time analysis, he added.

"The database will analyze long-term health care trends, but we also see cases where the analysis of real-time telemetry ... may indicate that someone's about to become more sick. It enables predictive and preventative care based on trends and on the more immediate data of patients to identify a high risk patient who should be monitored or attended to before a problem occurs," Spang said.

These systems are getting easier to use, according to IBM, and that should enable new discoveries in data. "That capability is part of Smart Analytics," Spang said. "It's about letting the system try to find the golden needle in the haystack."

It's also about people interacting with the system. "One client is expecting to learn things they don't even know by making the system available to a broader class of business users," Spang said. "The random human element is important to analytics as well."

He said the same thinking had been in play with Cognos, the business intelligence provider IBM snapped up in 2007.

"That's one of the values of Cognos: an easy and intuitive interface that enables flexibility and customization so that information can be put into its natural context and made available as a Web service," Spang said. "Information can be fed into other processes, and displayed in a portal."

While IBM has not yet disclosed pricing on the offering, it's aimed at being cheaper than the high-end, custom business intelligence enterprise deployments that now typify much of the space. Coupled with the lower complexity involved in deploying the solution, observers see Smart Analytics paying off.

"Cost and complexity ... are going to be the two things that business intelligence vendors are going to have to address," said Intelligent Solutions's Imhoff. "People are not going to be willing to pay what they paid in the past and ... we're now starting to see small to midsized businesses adopting business intelligence. They want the same sophistication [as enterprises.]"

"The BI vendors have got to address the problem of complexity much like IBM has done, making it as simple as little Lego blocks," she added. "Lowering both [cost and complexity] will remain the drivers in the future."

IBM Smart Analytics will be available on Sept. 29.

InternetNews.com Managing Editor Christopher Saunders contributed to this story.