Posted by: SJC
Business process automation, Database application, Database application front-end programming, Reporting, Software application development
Business Intelligence — I have heard it referred to as an oxy-moron, but I think I’ll not go there. Reading through the technical press one needs not look very far before coming upon writings about BI — implementation issues, how to’s, pitfalls and about BI project ROI. There is much ado about BI – but what about BI for the small business? Does it fit, and if so how?
In examining the question of how BI “fits” for a small company, I found myself asking the question “What makes BI – BI? How does it differ from reporting which may include graphs and really neat visual presentations similar in many ways to those of BI applications?” In researching this topic I found an excellent white paper entitled “Reporting vs. Business Intelligence“, written by Jim Hammond, President of RainMaker Software, Inc.
While the focus of this paper is BI as it applies to law firm needs and RainMaker’s software it provides insight into some of the key differences between reporting and BI which apply across all business applications.
Personally, I look at BI as being reporting on steroids. Mr. Hammond refers to BI as an “…ad-hoc way of easily selecting “exactly what you want” in the “exact format you want to see it in” type of reporting tool. I see reporting as presentation of the “details”, and BI as presentation of the “big picture”.
Enter the “Data Warehouse”. Mr. Hammond refers to the data warehouse as built “…for nothing more than precise, high-speed, data analysis…” While reporting generally is against a transactional database, the BI application runs against the “data warehouse” which generally will have summary information not readily available when reporting against the transactional system – hence the speed of data availability.
There are constant references to reporting against the transactional system as being “hard coded”, thus being limited by the design choices of the application developer. However, there is much consideration that needs to be made relative to the creation of the “data warehouse”. Will not the BI application require design of said “data warehouse”? To the best of my knowledge one doesn’t go to their friendly retailer and buy a “data warehouse” — perhaps an opportunity for new product?
As for small business situations, my experience is that in most there is no user who has the knowledge to access the “ad-hoc” capabilities of the BI system and produce meaningful, useful and accurate analysis data. The idea of “slice and dice” of data has much appeal, but my experience to date has been that users still consistently turned to the IT staff for their “slicing and dicing”, and the IT personnel ended up creating the equivalent of “canned reports” so that the users could just “point and click”. In this case what really resulted was reporting against the “data warehouse” information rather than the transactional database.
There is a cost associated with the creation of the “data warehouse” which can be prohibitive for the small business. The return on investment, however, can be significant. Getting into BI should not be taken lightly, as behind the “glitz” of presentation, its still really all about the data integrity driving the presentation – for any BI project to succeed one requires good data.