It’s a question, usually asked by an IT person, that Forrester Research’s Boris Evelson says he gets often: Boris, what do I put on the CFO’s dashboard?

“The CFO knows he needs one because everyone has one,” Evelson says of these conversations, “but he doesn’t know what to put on it.”  

That perennial business-IT question came up when Evelson and colleague Rob Karel recently looked for trends in the BI organizational structures of 197 companies. The survey is part of a larger report set to be published in the next few weeks on best practices for forming BI organizational structures.

The survey of “BI-savvy” IT professionals came up with one consistent, and familiar, theme: The business doesn’t take enough ownership of BI projects.

BI organizational structures tend to follow two extremes. They’re either completely siloed, and as Evelson puts it, the left arm doesn’t know what the right arm is doing. Or they are the other extreme—completely and totally centralized, and projects move slowly and nothing ever gets done.

Obviously, neither is the ideal. The best BI organizational structure lies somewhere in the middle. It’s a crucial point: All of BI’s best practices rest on developing the right BI organizational structure, Evelson says: The right structure for agility, flexibility and to be reactive.

And, predictably, these are also organizations where business takes proper ownership of BI.

A characteristic of a good BI organizational structure is one that separates data preparation and data usage. Forrester asked BI professionals who took the survey: “How closely coupled are your data preparation and data usage organizational structures?” A total of 35 percent answered that some are tightly coupled, others not; 24 percent said they are loosely integrated and coordinated; and 22 percent said that they are one and the same.

Business people shouldn’t be spending their time running batch jobs or putting the data into a data warehouse, Evelson says. IT needs to be in charge of those tasks.

But once the data is in one place, there is little reason why IT should be involved, he says. With all of the modern BI tools, any businessperson who knows how to use Excel can define what they want on their report, query and dashboard.

Another characteristic of the best organizational BI structure is one that separates the data needs of the front office and the back office. For example, accuracy is more important than speed to a CFO—a CFO can’t have incorrect numbers. But for the front office, it may be more important to be able to give a customer a timely, approximate answer. Knowing how to differentiate between the business requirements for accuracy and risk tolerance for latency is crucial, Evelson says.

“When IT starts feeding front office people the same approach as back office, that’s where the breakdown occurs,” he says.

Other interesting findings from Forrester’s survey include:

  • Cross-enterprise data usage is the main driver for BI support centralization. When asked how they assigned responsibilities of the centralized/shared BI support organization versus individual business unit support organizations, 45 percent said it depends on cross-data usage, followed by how mission critical the supporting business processes are, and how mission critical the data is.
  • Too few organizations take a quantitative approach for BI measurement. A total of 54 percent said they engaged in informal, qualitative measurements, while 33 percent said they didn’t have any real measurements at all.
  • The ratio of users to BI support staff varies widely. The largest percentage (16 percent) answered that there were 50-99 users for one BI analyst, architect and developer. But answers varied widely; ranging from fewer than 10 to 1, to more than 500 to 1.