How to Increase the Reliability of Your IT Infrastructure Using Predictive Analytics REGISTER >
Gartner describes several levels of BI deployment (see chart at left), from bread-and-butter management reporting to implementations of packaged solutions in isolated departments here and there, all the way to what it calls "BI enlightenment." At this level, information enables you to change the business and react faster than competitors to changing business conditions and new profit opportunitiesessentially, to innovate.
"What's been missing is strategic employment of BI," Gartner's Rayner says. "This goes beyond management reporting and transactional analysis. How is the business performing, how is it doing? How do we support the executives making decisions? Can we say, 'Here's where we are today,' and then forecast trends? You really start getting to enlightenment when you're pulling together data from a multiplicity of sources, from ERP, CRM and legacy systems. You increasingly need to work with data coming from outside the enterprisefrom trading partners or the supply chain. Only when you understand how all things fit together can you be enlightened."
Only a small number of companies, he says, understand the importance of BI at the strategic level and are attempting to use it in a comprehensive, integrated way. "Those who do so effectively will outperform their peers and win market share."
John McDermott is partner and practice leader for information management at Marakon Associates, a management consulting firm. The panoramic view from his 44th floor office in midtown Manhattan is symbolic of the strategic perspective his firm takes.
McDermott talks about the "information advantage" a company can gain if it understands the unique way its leaders make decisions and can deliver the information they need to make those decisionsbe it historical or forecasted, internal or external. BI software can help orchestrate the process.
"If you get it right, really get it right, you end up making better decisions faster than your competitors," he says.
And yet you can't turn the business over to a computer and turn off your brain. "There are certain kinds of data that don't lend themselves very well to automation," McDermott says. "For example, a deep analysis of competitor activity. People are able to see patterns in the data that we just can't possibly program into a system, even with some of the expert-based systems that exist today. It's much more practical to ask a person with good business intelligence to do it."
Comments on this story can be sent to firstname.lastname@example.org.