Good Can Be Better Than Perfect

Good BI is better than perfect BI.

We all know someone who's a maddening perfectionist. Over the summer, my family vacation included a visit with an architect cousin who's at long last putting the finishing touches on the "dream" house built for his mother. Considering how quickly other houses in the area have gone up, this one—well, let's just say that the years of delay have produced enough family turmoil to fill a novel.

Hugging the high ground over wildflower fields and a river, the house is wonderful. But it's not perfect. Tell my cousin that the tall living room windows look great and he'll grimace about what he really meant to have there. Contractors, masons, electricians: None among the diverse cast of characters did his or her job exactly right. "You have to learn to live with 80 percent of what you wanted them to do," he told me, "and that's when things are going right."

I was reminded of this tale as we were crafting our sixth annual issue on the state of business intelligence (BI). Research reports often describe the low penetration of BI tools into the potential user base. In her article, Cindi Howson cites such findings from The Data Warehouse Institute's (TDWI's) recent enterprise BI survey. The silver lining for the BI software industry, of course, is that there's room for massive growth.

Chances are, however, that the pace of expansion won't be as fast as BI partisans envision. Many experienced users of the technology view the BI Utopia as but a dream. Data quality will never be perfect. There will never be one single version of the truth. And not all decision makers will be using BI tools, much less a single, integrated enterprise suite.

BI's relationship with spreadsheets—long the bane of BI's existence—is a welcome example of a maturing industry. "Unlike [Microsoft] Excel-based spreadmarts, Excel can be a legitimate BI tool when used as a front end to an analytic server," states the TDWI report, written by Howson and TDWI director of research and services Wayne Eckerson. It may not be long before surveys lump some Excel use in with other BI tool implementation, which would certainly bump up those penetration figures in a hurry.

The truth is that "BI" didn't start with the invention of BI tools. The software has always had the task of breaking into a "mature" user base. Many organizations are just now reaching a tipping point where demand is loud enough that they're ready to commit to a strong BI infrastructure. "We reached a point where it was taking us a battalion of people to collate and send out homemade, paper-based reports to 5,000 users," says Steve Dewar, manager of Financial Systems at York University in Toronto. Dewar's group turned to BI tools from Actuate to deliver more timely data.

Key to getting the job done was, of course, Excel integration. Not only are users happier, but the University is finally getting full value out of its back-end Oracle/PeopleSoft applications. What's the moral of the story? When BI implementation stays focused on its mission of making the most of the data for better decision-making—and not on whether the user community is discarding old tools and buying in to the ultimate BI nirvana—BI "penetration" accelerates.

I'd like to close by welcoming new Change Agent columnist Bruce Silver, and also thank Don Tapscott for his many great columns as he departs to focus on other projects. Bruce is an expert in content and process management subjects, which are a key part of our coverage at Intelligent Enterprise. Let us know what you think of his column!

David Stodder is the Editorial Director and Editor in Chief of Intelligent Enterprise. Write to him at [email protected].