The New BI
The old BI sought pretty simple answers. How much of a particular product did we sell in this region in this time frame? How did this region compare to that region? If a manager was particularly BI-savvy, he or she might try multidimensional OLAP, in which you compare multiple factors with other factors in a complex matrix, cutting it into different views by this factor or that factor, “dicing and splicing,” as it was often referred to. At that point, however, the call usually went out for the Ph.D. statisticians to interpret the results.

Then there was the now apocryphal data mining story about diapers and beer, which drove data mining to become the BI rage for a while. Somebody looked at sales data, moved displays around, looked at the data again, and discovered that the store would sell more beer on Friday afternoon and Saturday mornings if it displayed the beer next to the diapers near the checkout counter. Eureka! Operational managers suddenly saw a tool that they could use to boost the performance of their business unit.

The new BI is moving beyond ad hoc querying, reporting, and even data mining as it makes a play for the executive suite. The latest BI hot button, aimed to grab the attention of a new generation of executives versed in information-driven decision-making, is analytics, the subject of Harris’s book. Even more than just analytics, the goal is predictive insights — the notion that you can model and analyze data to predict future behavior, such as how a group of customers will respond to a subtle change in product pricing and packaging.

“By analytics we mean the extensive use of data, statistical and quantitative analysis, explanatory and predictive models, and fact-based management to drive decisions and actions,” writes Harris. Predictive analytics are not exactly new. They have fueled lending processes such as credit scoring, in which a credit provider tries to predict the likelihood that a borrower will default. Financial institutions use it to predict churn in their customer base and to identify likely fraud. Harris takes analytics even further: “Analytics can support almost any business process,” she writes.

More than just analytics, executives need their BI to be forward-looking. “Now executives want a new BI that is predictive,” says Sanjay Sehgal, principal, CFO advisory services practice, Archstone Consulting. However, for most organizations, BI will end up delivering a blend of historical and predictive analysis as executives try to understand what has happened, what is happening now, and what they could change to get different results in the future.
 
In its own survey, InfoWorld magazine found that 62 percent of respondents report using predictive analysis now or plan to do so within the year. Of those, most were applying it to sales, finance, and marketing. Fewer were applying it to logistics/materials management or customer service/call centers.

Not only does the new BI have to be more predictive, but also it needs to be real-time. Relying only on data that is months, weeks, or even days old won’t cut it. “Companies need to move away from historical data to real time. They need the analytics embedded right into the business process,” says Molley.

Finally, to date, BI has focused primarily on structured data from databases and various applications. The next frontier for the new BI may be unstructured data. For example, what insights can executives glean by applying analytics to such unstructured data as call center or help desk records?

BI Obstacles
There remain a number of obstacles to BI, although technology isn’t one of them any longer. Rather, the obstacles today, surprisingly, continue to revolve around many of the same issues that have troubled BI since its earliest days: data availability, data quality, and agreement on the meaning of the data.

“The number one complaint I hear from executives is that the information doesn’t match,” says Sehgal.

This problem has been aggravated by the growth of the global economy. “With acquisitions, companies have grown globally. Now there are multiple instances of systems. Companies are pulling information from many more places,” Sehgal continues.

The solution is as obvious as it is old: “Companies must strive to get to one version of the truth,” says Sehgal, a refrain consultants have been repeating since BI began. Yet even now, companies have difficulty agreeing on such basic numbers as what constitutes sales revenue. Is anyone surprised that often the numbers don’t match?

Organizational silos continue to frustrate BI efforts. “Companies have to connect the dots across all functions,” says Molley. C-level executives will continue to remain in the dark as long as the data they need sits in each individual department. Or, they will have to spend a fortune, as Songnefest points out, to pry that data out of each department and normalize it before they can learn anything from it.

While the trend toward information-driven BI analytics in the executive suite, propelled by books like Harris’s, appears inevitable, there remain a large number of diehard decision makers who prefer the from-the-gut approach. Downplaying analytics, these executives will make their decisions in an instant based on their gut instincts.

These from-the-gut decision-makers have found support in Malcolm Gladwell’s book titled Blink (Little Brown, 2005), which promotes decision concepts around what Gladwell refers to as the adaptive unconscious. His message: “Decisions made very quickly can be every bit as good as decisions made cautiously and deliberately.”

The two approaches, Harris’s and Gladwell’s, may not be as contradictory as they initially appear. Actions taken in a blink, it turns out, often are based on years of experience that had been thoroughly digested before that seemingly instantaneous decision.

If so, BI and analytics in the executive suite may simply be a way to systematize and leverage the aggregation and digestion of the organization’s collective, adaptive, unconscious experience.