Let there be light

Illuminating the path to insightful business decisions can cost a fortune, but what if executives do not know what they are asking for?

Cutting the ribbons off a shiny new IT package is usually cause for celebration in an enterprise, though the rise of cost-conscious attitudes is seeing much less red linen go under the knife. Business Intelligence (BI) is no stranger to these types of reactions, as many players in the IT world still do not understand the technology or - perhaps more importantly - do not see how such an expensive piece of software is going to result in a clear return on investment.

When the price tag has as many zeros after it as a BI implementation does - not to mention the cost of setting up a data warehouse to feed your workhorse of choice - it is no surprise that some CIOs eye the technology with a large dose of caution. That is only one segment of the market however, with an increasing number of senior managers responsible for all things that go whir and clunk in an enterprise, becoming almost evangelistic about the benefits of BI.

There also exists a culture of misinformation about BI, possibly due to the fact that the term itself suggests an ambiguous definition of any intelligence related to the business. While there are undoubtedly some well-paid marketing executives who can spin the term with glee, the truth is that BI is defined differently by various organisations and even vendors like to argue about the semantics of what the label means.

James Richardson, research director at analyst firm, Gartner, explains that BI includes online analytics processing, data mining, predictive modeling and advanced visualisation.

"They are a part of BI and they have been a part of our definition for a number of years now, though there is an issue with that to be perfectly honest. They are the least-used parts of BI platforms. When we surveyed part of the magic quadrant that are using BI we asked them which pieces they are using now - across the vendors they are using - and the reality is most BI today is still being used for historic reporting with dashboards, so people are kind of driving using their rear-view mirror," he says.

While the analysts and the vendors usually agree on the standard definition for a technology, when it comes to BI all bets are off because there are no accepted industry-wide parameters for the software. This has come to a head of late with vendors like SAS pushing for a differentiation between BI and business analytics, which is generally considered to be part of BI, as can be seen in the Gartner definition.

"BI has two definitions in the market, there is the small definition of BI and there is the broad definition. The small definition is querying and reporting - so reports being churned out by some sort of front-end system. The broad definition is a process that enables you to get intelligence from a technology system, now the broad definition is what SAS subscribes too," says Carel Badenhorst, head of technology practice at SAS in the Middle East and Africa.

Badenhorst explains that the way SAS sees it, BI is basically looking at historical data while business analytics is looking at forward analysis. The only problem with this approach is that it doesn't account for the fact that all analysis is based on data, and a large chunk of that data is generated from events that have occurred in the past.

For Peter Thomas, independent consultant and renowned internet blogger on BI, the push being made by vendors like SAS to make business analytics its own field is just a marketing manoeuvre.

"It's entirely marketing; there has been an awful lot of consolidation with the main BI players recently. Cognos folded into IBM, Business Objects going to SAP, Hyperion going to Oracle, so SAS is the last large independent BI vendor," he says.

"They are looking for how to differentiate themselves; they have always been a bit different because their tools - as well as being BI-focused - are used by analytical professionals such as insurance actuaries, investment bankers and for some of the number-crunchers on the back end. It is quite typical that you will have something like SAS in-house for specialists to use alongside a broader, base BI tool like a Cognos or what have you. SAS has had a little bit of differentiation and the area where they have had the greatest strength is in the analytics space, which is part of the overall BI landscape in my opinion," adds Thomas.

At the end of the day, a definition is only as good as those who believe in it, so if an enterprise chooses to define BI as something entirely different then there is nothing wrong with that. Indeed, there are many organisations that don't subscribe to the vendor-definition of BI and instead look at the software as BI applications, with BI describing the larger process of dissecting and analysing data in report format. The term itself - business intelligence - may have been coined by Howard Dresner in 1989, but the technology has evolved considerably since then, and what organisations take it to mean has changed with it.

Erwin Bamps, CEO of GulfCraft, believes that BI applications are about collecting data out of your systems, whether it's something from your supply chain, personnel, manufacturing or other activities like sales and marketing and presenting that information in a way that dissects business-critical operations.

READERS' COMMENTS

MORE FROM ARABIANBUSINESS.COM