Getting Smart About Business IntelligenceBy Michael Vizard | Posted 2008-01-30 Email Print
The evolution of enterprise-class apps and services will end the generation of useless reports.
For almost as long as it has existed, business intelligence has been one of the applications on the cusp of going mainstream. After all, IT should be about a lot more than collecting data. We need data that is actionable.
Collecting data is a lot easier than creating useful information from all the data collected. As a result, IT people have been charged with creating the reports because they understand the intricacies of the data warehouses and applications. But, of course, no one on the business side is happy with the way the reports are tabulated and formatted, because they’re created by IT people who really didn’t understand what they were looking for in the first place.
Truth be told, the responsibility for generating these types of reports should rest with the business side. For example, Southwestern Energy has created a special business intelligence team, led by Chad Erman, which crafts the company’s BI strategy in concert with IT. This means business users typically get what they really want and shorthanded IT departments don’t have to dedicate people to running reports that are ultimately discarded.
As this trend takes hold because the business intelligence tools are becoming more accessible to the average user, the number of users taking advantage of the tools will rise. This presents a bit of a challenge to the vendors in this space because their collective cornucopia of products and modules dissuades people from invoking business intelligence applications that make it too hard to figure out which function or toolset will accomplish the desired task. For the enterprise, the challenge is figuring out if you can afford the applications, and how many licenses you need to support your users.
The vendors will argue that this modular approach is needed because customers don’t want to pay for things they don’t use and this allows them to see what they’re actually using. But vendors also abuse this strategy to milk more revenue from the customer—more than a few of these modules are really features of some other product masquerading as a separate necessity. You can, of course, get around this by buying an enterprisewide site license, but that will cost you a fair amount up front.
The good news is that, with the acquisition of Cognos by IBM and Business Objects by SAP, there may be a lot more willingness to explore the constraints holding back the adoption of business intelligence. After all, these vendors have to recoup their investments by driving up the number of licenses being used. At the same time, the consolidation in this space is also likely to attract interest from Microsoft, who probably sees business intelligence as the next great natural extension of the Microsoft Office platform. We should also expect to see a rise in software-as-a-service offerings that target the business intelligence space, especially for more routine reports that don’t require a whole lot of customization.
As Cognos president and CEO Rob Ashe put it recently, what we’ve just been through is “a 10-year beta cycle” for business intelligence applications that ultimately will prove a bigger market than the transaction processing applications typically dominating IT today. And those next-generation business intelligence applications are going to look a lot more like Web 2.0 consumer applications than the canned reports most people see today
For example, envision a world where not only is search embedded in every application, but any handheld device can be used to kick off any type of mash-up report on demand because the back-end architecture of the business intelligence application relies on a service-oriented architecture. Better still, imagine being able to reverse engineer a particular workflow or business process simply by tracking the metadata that sits above the business intelligence framework. No more guessing about who’s using what application and why.
With any luck, instead of waiting another 10 years for business intelligence to become really useful, this might all play out in the next two years—maybe sooner. Southwestern says the approach works just fine for them.