LESSON - There’s More to Business Intelligence than Just Software
Without the input of business users, your BI solution will fail.
By Gabhan Berry, CTO, XLCubed, Ltd.
Business intelligence is all over the news these days. It seems that not a week goes by without a business intelligence–related headline being splashed across the industry press. As the hype grows, the stakes rise, and the anticipation for businesses to get “real intelligence” from their solutions mounts. Companies buy the best software they can afford, build it as quickly as they can, and expect the insights and intelligence to jump out onto their screens… and why not? Isn’t that what every vendor promises? Business intelligence should deliver business insights, not data.
But there is more to good business intelligence than good software.
There is more to delivering a business intelligence solution than having IT departments build a few analysis cubes, purchase a reporting tool, and hand it over to end users. Designing and building a business intelligence solution is not overly complicated, but it does require some up-front thought and planning. With more and more business intelligence vendors proclaiming simplicity and ease of use, fewer and fewer business intelligence solutions seem to deliver actionable insights to the business.
Time and again, I have seen problematic business intelligence solutions. Sometimes the problems are related to failing software, but more often the cause is human error: human error in the form of poor business modeling.
In order for a business intelligence solution to discover insights and allow the business to act on its data intelligently, the solution has to model the business problem accurately. This seemingly obvious statement is all too often overlooked or not even considered. How can a model that does not reflect the real world be of any use to people in the real world? How can a model that does not understand the problem be used to solve the problem?
In order for a business intelligencesolution to discover insights andallow the business to act on its data intelligently, the solution has to modelthe business problem accurately.
Before designing and building a business intelligence solution, the business needs to take time to understand the exact problem they are setting out to solve. Next, they need to get the right people around a table to provide an accurate model of the business, and it is here that the biggest mistake is made.
The right people for modeling the business are not IT staff; they are the business analysts. Most IT staff are too far removed from the day-to-day running of the business to be able to build an accurate model of the business.
Business analysts need to be involved in a business intelligence solution. It is their job to use the solution to enable the business to act intelligently on its data; they are the users of the system. Airplanes are not designed without pilot involvement; racing cars are not designed without driver involvement; so why would you build a business model without the involvement of a business analyst?
And yet so many do.
A failing business intelligence project is one that segments the IT staff from the business staff. Different groups of people with different objectives are rarely able to come together as one. The IT department may build “technically” good multidimensional analysis cubes, but they are not useful if they do not reflect the business accurately.
Business intelligence is about uncovering insights. Important bits of information that impact your business are buried within the data. Data is critical in business intelligence, but data alone is not enough. You need a good model, and that model is going to come from the business analysts. Once you have a proper model, then you can begin to design, build, and deliver a business intelligence solution that will deliver information and insights to your desktop!
This article originally appeared in the issue of .