News Stay informed about the latest enterprise technology news and product updates.

The upcoming revolution in business intelligence systems

A revolution is coming to business intelligence systems that's fueled by more compute power, mobility, data from everywhere and the adoption of predictive analytics.

Business intelligence (BI) systems are evolving so rapidly as to be almost unrecognizable. Just a few years ago, BI meant clunky, backward-looking financial analysis systems and interfaces deemed unfriendly by users at large. Improved computing power is increasing by orders of magnitude the amount of data that can be crunched, while the advent of new types of media is changing the kinds of data to be crunched. At the same time, the rise in mobile business computing has begun to alter how BI applications are being delivered.

CIOs forging their BI strategies soon will have to grapple with business intelligence systems that are little more than buzzwords today: social network analysis, organizational network analysis, context-aware computing, even sentiment analysis. Sounds far out? The BI division at Tata Consultancy Services Ltd. is developing software that instantaneously analyzes shoppers' facial expressions (confused, angry or bored, for example) caught on in-store video cameras, and alerts salespeople in real time to take appropriate action.

If industry experts are correct, next-generation business intelligence systems -- or business intelligence technology 3.0, as some call it -- will be a heady mix of internally generated business data correlated in the blink of an eye with real-time intelligence culled from multiple sources, such as social media, newsfeeds, video and sensor-enabled environments.

Not only will next-generation business intelligence systems solve the latency issues that dog many classic BI operations, but business intelligence technology 3.0 will use predictive analytics routinely, conjuring what-if scenarios to inform real-time decision making. Complex event processing, the detection of patterns lurking in petabytes of data, will no longer be the sole domain of the global investment banks or the military. Every time a customer asks about a product, a company will use powerful analytics software to predict how that customer would be served best at that moment, based on past preferences and the current context. In fact, Gartner Inc. recently predicted (presumably by using predictive analytics) that by 2016, all enterprises will be using next-generation analytics to forecast the future.

"But [these next-generation BI advances don't] mean, run out and immediately update and add a bunch of compute capacity. It means, be ready," advised Carl Claunch, a distinguished analyst at Gartner, during the consultancy's symposium event in October.

Business intelligence systems for business value

CIOs also need to be ready for the explosion in mobile business intelligence applications, said Howard Dresner, an independent BI analyst who surveyed 200 user organizations about their mobile BI strategies. Although Blackberries and other smartphones have hosted mobile BI apps for years, their small screen has been a roadblock to using them, he said. The iPad and other tablets overcome this problem, and his research shows that CIOs should expect enormous growth in mobile BI. "There is no longer a reason not to know something," he added. Security remains an issue, and organizations will have to stay vigilant about protecting their data and have processes in place to "wipe" a device if it is lost or stolen.

At the moment, however, companies' main challenge in this revolution in business intelligence systems is figuring out which kinds of BI tools will give them more business value, rather than just more information, analysts and CIOs say.

Companies will discover over the next few years where new business intelligence systems, such as those for predictive analytics, can be applied to provide value for their industries. "That is when you invest," Gartner's Claunch advised.

The 'uncharted territory' of business intelligence systems: Predictive analytics

Atefeh Riazi, CIO of the New York City Housing Authority, is identifying already where predictive analytics can improve the agency's mission of providing affordable housing to poor and moderate-income families.

A simple example: As part of a long-term initiative to figure out which conditions actually improve residents' quality of life, Riazi's team is studying the impact of security cameras on violent crime. The assumption is that cameras prevent crime. However, after running correlations on a decade's worth of information from multiple sources including police reports, the team found that cameras deter vandalism but have no effect on violent crime after they've been in place for two months, unless they're coupled with other factors, including an effective intercom system and random patrols by police officers. That information will help direct how the authority invests in security.

Riazi described running similar scenarios to figure out which of the authority's buildings are most likely to be affected by flooding after various amounts of rainfall, so action can be taken to stave off damage to boilers and elevators.

Analytics will take us to uncharted territory. A lot of what we assume, what I call urban legend, is going to be proved wrong.

Atefeh Riazi, CIO, New York City Housing Authority

As for technology, Riazi's team uses a variety of BI tools, from SAP AG's BusinessObjects and IBM's Cognos and WebSphere to Visokio's Omniscope, a nifty application on Riazi's desktop that runs correlations on large data sets. The products, she insists, are not the point, however. "The technology forces us to think more analytically and use information to predict what will happen so we can be proactive," Riazi said. Finding the people with the skills necessary to do analytics is difficult; convincing customers of the results generated by fact-based analysis can be even harder. "We are taking baby steps," she said.

Even more complex, Riazi adds, will be the business intelligence projects aimed at learning which conditions -- from proximity to transportation, supermarkets and houses of worship, to more intangibles, such as close-by and close-knit family -- really do help the authority's customers lead happier lives.

"Analytics will take us to uncharted territory," Riazi said. "A lot of what we assume, what I call urban legend, is going to be proved wrong. When you start proving urban legends wrong, you will shift investments to get better returns, and when you do that, you will cause some disruption in the business and among customers."

Wayne Eckerson, director of research at TDWI Research in Hingham, Mass., and author of TDWI's BI Maturity Model, agrees that CIOs' use of predictive analytics will be dictated by their ability to find good analysts -- as well as by their companies' culture. People who are good at this kind of business intelligence work are not only critical thinkers but also "doggedly persistent," he said. They like to experiment. They know the data and the business processes, and they understand how to use the tools -- a rare combination. And they are pricey, he said. Candidates for the job are as likely to come from MBA programs or the social sciences as from more technical backgrounds. Nevertheless, luring the best analysts will make no difference to a predictive analytics program if business executives are not willing to test assumptions, he added. In fact, if business leaders are not "'show me the data' types," Eckerson said, CIOs may be out of luck.

Let us know what you think about the story; email Linda Tucci, Senior News Writer.

Dig Deeper on Enterprise business intelligence software and big data

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.