I'm mildy obsessive. When I get interested in a topic, I sometimes find it tough to let it go. I'm one of those folks who gets a little too interested in cocktail party chatter and ends up monopolizing a fellow guest with question after question about his or her work, hobby or what have you. My obsessions are "mild,"  so I won't follow the partygoer home or anything. However, this trait -- along with a dearth of other marketable skills -- had a lot to do with my becoming a journalist.

So when I write a fairly detailed story about an interesting topic, I am often reluctant to move on. And thanks to this blog and my reading habits that support it, sometimes I don't have to. A few days after I wrapped a story on business intelligence initiatives, I saw a TechRepublic column written by Scott Lowe, the CIO for Missouri's Westminster College, in which he reiterates several points made in my story.

The biggest takeaway, both from my story and his account of the college's experiences: Get business users involved. While that's important with most technology-driven projects, it's especially true for BI since at least some users rely on it to get their jobs done in a way they never will for, say, virtualization. While virtualization obviously offers business value, end users care little about the particulars. In fact, it'll be transparent to most of them.

Though Lowe suggests including everyone from the CEO on down, he makes it clear who he thinks ultimately determines the success of a BI project. He writes:

It’s the responsibility of the CIO to guide the project, make sure that what’s being requested is technically feasible and, in many cases, to assist stakeholders, including the CEO and department heads, to figure out what they want and what their part is in the BI implementation.

He describes the CIO as "a facilitator, enabler, and advisor in BI efforts."

Lowe also echoes a point made by one of my story sources, independent BI consultant Peter Thomas, who wondered why any organization would undertake a BI initiative unless it was prepared to make some fairly major accompanying process changes. One of the four "keys to success" that Lowe includes with his column:

The organization must be willing to appropriately react based on the information gleaned from the BI system. If, for example, data begins to clearly point to a drying up of the market, the organization needs to be willing to not stick its head in the sand. If this is the reaction, why use BI at all?