It can get hard to discern what makes agile business intelligence effective, with vendors insisting that flexibility, do-it-yourself BI for end users, and hiring a few scrum developers "automagically" delivers agile BI. However, the secret to effective business intelligence is fundamentally very simple: It’s effective BI because it’s truly agile, and it’s effective agile BI because it modifies agile practices to the needs of BI – just as a truly agile process should.
The five habits of effective agile business intelligence implementations follow from this principle. Here they are:
Habit One: Connect to the customer
The customer is not usually a company customer or prospect, but rather a corporate executive or business analyst. Without their buy-in and frequent input (or that of a proxy, such as a consultant) as to what matters most to them, BI "agile" efforts can get badly off track, ruining the reputation not only of the BI implementation but of "agile" in general.
Habit Two: Get buy-in for agile metrics
If you are judged by cost, there will be constant confusion as to how to assign costs to projects whose targets and accomplishments frequently change. If possible, the metric should be along the lines of time to customer satisfaction or time to value.
Habit Three: Get agile IT involved
The new trend for DevOps – finally coordinating development and app administration according to agile principles – is especially important in making sure that the deployed business intelligence solution continues to deliver user satisfaction rather than being slow or unavailable. It also ensures that IT gets the solution out there quickly.
Habit Four: Apply agile database design
Use the database analog of "refactoring" – creating schemas for the data that are easily changed – when determining the best querying schemas for the solution.
Habit Five: Virtualize the BI solution
I realize that virtualization is an over-used word that suffers from marketing hype. In this case, however, it means specifically that the BI solution’s "view" of the data should not be tied to one database or even one vendor database. Nor should the queries be involved. There’s a simple way to do this: Use a data virtualization solution from a vendor such as Composite Software, Denodo, Informatica or IBM (WebSphere Federation Server).
The list is short and sweet. But then that’s the way with agile, isn’t it? And with agile BI, the rewards go straight to the top and bottom lines.
Wayne Kernochan is the president of Infostructure Associates, an affiliate of Valley View Ventures that aims to identify ways for businesses to leverage information for innovation and competitive advantage. Wayne has been an IT industry analyst for 22 years. During that time, he has focused on analytics, databases, development tools and middleware, and ways to measure their effectiveness, such as TCO, ROI, and agility measures. He has worked for respected firms such as Yankee Group, Aberdeen Group and Illuminata, and has helped craft marketing strategies based on competitive intelligence for vendors ranging from Progress Software to IBM.