DON'T SHOOT THE REPORT WRITER

For the past 10 to 15 years or so, the BI industry has been conceived and built around one basic premise: to empower business users by reducing or eliminating their dependence upon IT resources to access the information necessary to understand and operate their business. In short, to help make end users self-sufficient and self-servicing. One of the primary solutions to spring out of this was of course the whole concept of the data warehouse/data mart.

When the idea first emerged the industry hailed the concept of the data warehouse as the “final solution” to the pent up demands and frustration of millions of business end users who for years had been pleading with IT to provide some way, any way, for them to get quick and easy access to mission critical information. Users were told that the solution to their problems had finally arrived. However, it soon became apparent that this new way of doing things would produce rather mixed results at best. As Wayne Eckerson and Cindy Howson state in their article, “Moving BI to the Enterprise”, “In reality, however, self-service BI proved overwhelming for all but the most sophisticated power users.”

In my own experience in working closely with business end users for the past 20 years, I encountered this reality gap first hand. I have trained and mentored hundreds of end users using a wide variety of user interfaces into many types of both custom and off-the-shelf applications, including spending the past four years focused on assisting business users in accessing data in an SAP BW warehouse. Something I have seen time and time again is an overestimation on the part of IT decision makers of the capabilities of their end users. We seem to think that if we can simply clean up the data and pull it together in one place then the users can just dive in and find what they need and everyone will be happy. If my own experience is any indicator, this has proven true for a select few, but not so for most.

Before I continue, I need to make one thing very clear. Even though my own experience is on the front end (report development and delivery), I know enough of what is required on the back end of the process to appreciate what kind of effort it takes to get disparate and often times ugly data into a form and place where it can be rendered usable to the user. In fact, I often compare the reporting process as the “tip of the iceberg” when reporting off a data warehouse. It's the part everybody sees and can seem quite impressive until you go beneath the surface and see the rest of the story. Trust me: I have a great appreciation for the skill and downright determination required to assemble a properly designed data warehouse.

That said, just as the tip of the iceberg is the only part the ordinary mortal ever sees, the same can be said of the reporting interface that is used to look back into the warehouse: it's all the user knows about or cares about. In fact, nothing can be more frustrating for a user than to know that their IT department just spent many months (if not years) carefully designing and constructing a data warehouse only to find out that it's not quite as simple to get the data out as they may have been led to believe. So while the reporting interface may indeed be the tip of the iceberg, let's all keep in mind what happened to the Titantic.

One of the more recent developments in BI has been the increasing number and popularity of what I would call “intelligent” reporting interfaces. The grand-daddies of them all were Executive Information Systems (EIS) and Decision Support Systems (DSS). These have since been transformed over the years into more sophisticated interfaces like dashboards, scorecards, and KPI tools. While I am personally excited about the potential of these more intelligent interfaces for providing insight into critical business performance factors, I don't think anyone would suggest they present a complete solution to the user community at large. In my mind, it is because of one simple fact: the devil is in the details.

On of the unfortunate misconceptions about these high level “executive” interfaces is that decision makers would rather not be bothered by low-level details. In an article entitled “Reality IT: A Big Myth About IT Needs” Gabriel Fuchs states the following: “I have yet to see a senior manager that is content with an overview of the business activities - unless he or she is incompetent. Any and every competent senior manager that I have met always wants detailed data for specific operations - and often on a regular basis. Dashboards are great for management meetings, but whoever runs a company based on this is not running a company for long.”

So let's not be so quick to write off the report writer. It would be a mistake to assume that these new interfaces, while quite useful in their own right, will somehow negate the need for on-going operational (or detail) reports. My own feeling is that there will be a need for some time to come for individuals in any given organization who possess good report writing skills. And in an SAP BW environment it is critical that your report writers utilize a report (or content) development tool that provides for the greatest flexibility in information presentation options and integrates as seamlessly as possible into your existing landscape (utilizing existing InfoProviders and security models). In a future installment I will discuss some of the leading report writing tools and look at how well they do on both of these fronts.

By Mike Garrett , Data Management Group Senior Consultant