Levels of business intelligence (page 1 of 2)

By Jim Mortleman

It may be a tongue-in-cheek comment, but reflects the reality that where you are is not down to the tools you have, as much as who's using them and how. Clive Longbottom, principal analyst at Quocirca, says: "BI is greying. It no longer makes sense to say there's one set of tools at this level, then a big jump to another set of tools at the next level."

Even standard productivity applications like Microsoft Office today provide basic BI features. "For example, in Office 2007, you can create pivot tables in Excel very easily. And the next release will include a raft of new features such as graphical trending, which would previously have required dedicated BI tools," says Longbottom.

"These in-built tools are very powerful in the right hands - and harnessing that power is down to user training not technical skills."

The next level of sophistication is the ability to analyse real-time data. "That could reside in Excel spreadsheets, although it will more likely be in an Oracle or IBM database. Companies need something that sits on top of these data sources, maybe pulling several together, and allows users to examine what's happening very easily, often graphically," says Longbottom.

"Here you'll be looking at companies like Tibco Spotfire, Panopticon and Information Builders. It's also the direction Oracle's taking Hyperion, IBM is taking Cognos and to a lesser extent SAP is taking Business Objects."

Today, you also find user-friendly BI tools built into applications such as CRM and ERP. "There are tools included to see what particular customers or groups are buying, who has been buying a particular type of product and so on," says Longbottom.

"Then there are web-based analytical systems such as Google Analytics - which tell you things like how many people have come to the site, what tracks they took to get there and so on. This is powerful stuff - not only for CRM but for marketing and sales, as well as for feeding back into inventory management and manufacturing."

At the higher end, you have more technically advanced systems, aimed at those companies with massive datasets that do not necessarily know what information they are seeking. At that level, businesses need the system to present information in different ways.

"For example, in retail the system might look at various factors to ascertain, particularly with perishable goods, that you put the right quantity of the right goods in the right place at the right time," says Longbottom. Such systems, he says, require the business has considerable expertise - both among those creating the data and those in the IT department implementing and managing the system.

Oracle's business development director for EMEA, Jon Ainsworth, agrees talking tools is no longer the best way to consider BI. "You should be thinking about how they work together," he says, although he offers a quick primer for the confused.

"In terms of sophistication, the simplest tool is vanilla enterprise reporting. Next up is dashboarding technology which is more interactive and dynamic, showing standard KPIs [key performance indicators] and so on in a simple graphical view. More recently, such tools have added features like alerting and mobile access," he says.

"Beyond that, you have ad hoc reporting. This is more self service. If the user isn't getting the information they need, they can roll up their sleeves and dig around without having to rely on IT.