BI Megatrends

Richer, Actionable Analytics Stoke BI Demand

What defines BI success? "Keeping up the momentum: You have to continuously address the needs of the data curious," says Jody Porrazzo, director of econometric risk strategy at Apex Management Group, a consulting and insurance provider that specializes in actuarial, underwriting, claims management and auditing. "People begin with static reports but then start to see relationships, want to know more and want to do something," Porrazzo says. "For example, I want to identify all the risk factors for heart attacks. I can't contact the actual patient — I'm an econometric risk strategist. But if I find something, I want to get that information right away to our clinical people, who could get those at risk into the right care program. A person's life might be saved. That's ROI."

Apex is riding strong demand for rich analytics from a variety of health-care concerns. The company uses SAS Enterprise BI Server and data mining tools to understand the complex relationships between different plan offerings, disability benefits, medical system utilization — and, of course, cost. Porrazzo sees text mining as the way of the future. "What's it going to tell me? Best practices and outcomes," Porrazzo says. "Imagine if you had a data set, organized it according to the major disease categories and then analyzed the data with a text mining tool. You'd get a much clearer picture of what really produces the best outcomes."

Supplying richer analytics with less delay requires changes to the BI infrastructure. One future step for many organizations is adopting 64-bit processing platforms, which allow complex analytic queries to be dealt with in memory. "We never hit disk," says Porrazzo — Apex runs on Hewlett-Packard hardware with Intel Itanium 2 processors. "A lot of memory is important if you know anything about health-care data. I can go through 17 million rows for just one quarterly period." BI must keep pace with inquiring minds.

Technology Innovation Breaks the BI Mold

Companies need BI to be more than batch-oriented query and reporting. Data warehousing and integration infrastructure choices are opening up, especially with the maturation of federated enterprise information integration systems. But are BI tools breaking the mold?

While the major suites incorporate important features as they go mainstream, true innovation, as always, comes from small vendors with fresh approaches. Qliktech is anticipating the platform shift to 64-bit computing and larger and more affordable memory. Its technology does its number-crunching in memory. "We don't force users to live with prearranged cubes, aggregations and views," says Anthony Deighton, Qliktech's vice president of marketing. "It took 32-bit processing to open the way for multitasking. We see 64-bit processing as that kind of fundamental shift, with BI as the killer app. More and more organizations want to do things that are physically impossible on 32-bit platforms."

A good example might be a financial services organization that needs to do complex trend analysis on real-time data feeds to develop investment offerings. Canned reports filled with stale data don't cut it; and, unfortunately, current alternatives are expensive.

Another innovation is a change in the relationship between transaction systems and BI. Usually, BI comes after the fact. "Real-time" BI notions typically focus on speeding up data capture or delivery but not the analysis itself. Edge Dynamics offers technology that combines transaction processing with "inline" analytics: the ability to analyze transactions before they're recorded into core ERP systems.

Edge Dynamics is responding to the heat on pharmaceutical manufacturers that are working with numerous wholesalers and levels of distribution. "There might be thousands of distribution points to which orders are being shipped," says Edge Dynamics CEO John McGrory. "The manufacturer needs to keep close control because waste is very expensive and has other negative effects on business relationships. If you think about getting in front of that order stream, there are an awful lot of ways you could take corrective action before it's too late. You could make sure that every order that comes in is the perfect order."

Edge Dynamics' customers are keeping quiet because they fear Wall Street and regulators might hammer them for even admitting that they're correcting inventory and distribution problems. That's when you know BI is at its most innovative: No one wants to talk about it.

Sensor and Location Data Sharpen Intelligence

Thanks to the increasing adoption of XML, unstructured content — including images, e-mail, forms and documents — are coming into range for business analysis. Business Objects, Cognos, Information Builders, Hyperion, SAS and other BI tool vendors are letting customers leverage unstructured data largely through technology partnerships with search, text mining and other specialized analytics providers.

The BI community, however, has been slow to react to the rising interest in and deployment of sensors, radio frequency identification (RFID) and location technology, which includes maps and geographical information systems (GIS) data. Data volumes from these sources are large and growing, but the analysis requirements may not yet be defined to the point where BI vendors can market commercial tools. User organizations are left to develop their own analytics or work with the vendors' services. (See "Beyond Quick-and-Dirty RFID")

Location technology may be a different story, as Seth Grimes discusses in his column. Cox Communications, a large U.S. cable provider, uses integrated technology from MapInfo and MicroStrategy to manage and analyze customer, fiber and service location as well as other operational data. Trucking and logistics companies are applying predictive analytics to location data to improve safety and asset management. Retail, insurance, energy and law enforcement concerns are integrating location intelligence with other data, often with the help of advanced data visualization.

As computing and communications converge, data becomes less static. BI must be in motion as well.

What's Next in BI products

Putting all six megatrends together, what do we see? First, maturity: While BI tool providers are consolidating their suites to address mainstream needs, IT is zeroing in on the infrastructure problems that most threaten BI's growth and success. Second, a new wave of technology is on the way that could help BI answer the call for real-time, actionable analysis--perhaps just in time to field those queries coming from the "masses" working on operational performance problems.

And, finally, we see excitement: as Cindi Howson describes in the second part of our cover package, BI "ease of use" is becoming more than an oxymoron. There's still a way to go, but BI is gaining ground as a must-have technology for business decision-making.

Business Intelligence Business Intelligence

Business intelligence (BI) is more than a suite of software tools: It's an essential and continuously expanding activity focused on accessing, analyzing and acting on data to meet business objectives. Today's BI tool market is boiling down to a smaller number of players offering comprehensive suites that deliver mainstream functionality. However, the unending search for data-driven insight continues to draw in new, more powerful technology as well as creative approaches to applying knowledge to improve business processes and performance.

Major trends in BI implementation involve front-end tools and back-end infrastructure. On the front end, browser-based BI and dashboards are evolving the view of information away from standard reports toward metrics and key performance indicators. Spreadsheets are the object of intense interest as BI vendors compete to supply the fullest integration so that nontechnical users can leverage all data sources. Specialized tools for data visualization are taking graphical analysis to new dimensions.

On the back end, regulatory compliance puts pressure on IT to ensure high data quality. Strategic business objectives are forcing organizations to break down information silos so enterprise BI and performance management tools can gain a single view of customers or other objects of interest. Better infrastructure will lower BI costs and let organizations serve the right information at the right time.

Our cover Package continues on with Cindi Howson's "Flying High with BI," a look at the maturing BI market.