IT obviously plays an essential role in every technology implementation. But I think any implementation must begin with an examination of the business drivers and goals. Otherwise, companies risk letting IT personnel rule out potential solutions based on their own ingrained tech preferences.
That was my initial thought after reading a Dashboard Insight piece written by Lyndsay Wise, president of WiseAnalytics, whom I've interviewed several times. The article, which offers suggestions for companies evaluating business intelligence dashboards, is packed with good advice. And yet, I take issue with the way Wise presents her five primary considerations. That's because she starts with IT integration. Her reasoning is sound. She writes:
In most instances, a current infrastructure exists whereby certain solutions will be easier to maintain and manage, thereby lowering the effort of IT staff and enabling them to provide better service to the business. IT people take into account the required and desired features, overall interoperability, customization versus what comes out of the box, integration, overall maintenance, security, etc. when conducting their evaluations.
While it's true software that requires an undue amount of customization, integration heavy lifting or other excessive intervention from IT likely won't be a good choice for business users or IT, I do think many IT departments simply prefer to go with known quantities rather than considering new options. (Its the "we've always done it this way" syndrome.)
So IT provides business users with a short, pre-vetted list of "acceptable" software. Yes, it makes IT's job easier -- at least at first -- but there's a risk that solutions that best fit the business needs simply won't be considered. This tendency has contributed to the trend of business users quietly adopting solutions not sanctioned by IT and hoping IT doesn't find out about it and bust their chops.
Wise does imply IT should bring business users in at this stage, writing "cohesion between business and IT ensures the most accurate and relevant solution choice to the business." When I e-mailed Wise and asked why she listed her considerations in this order, with IT first, she responded:
The reason I listed IT first is because in many BI situations, even though projects are slowly moving toward being driven by the business, the reality is that many IT departments still play a big role in designing and maintaining the infrastructure and still aren't fully open to the idea of having projects driven by the business unit.
This is a huge issue, not just with BI but with any technology project. Wise mentioned she tends to get more support from IT by putting its considerations first. I wonder if other consultants and BI vendors experience this? I suspect so. On Friday, I wrote a post in which I cited remarks from Roman Stanek, CEO and founder of GoodData, a provider of cloud-based business intelligence software. He said, in part:
BI, even though it has "business" in its name it is actually fundamentally an IT initiative and I don't think that IT departments know how to solve business problems.
Wise also told me "there are many solutions that can be implemented on the business side independently of IT." That's certainly true, and this is sometimes used as a selling point by vendors using a cloud-based delivery model, like GoodData. I'd agree that cloud-based BI works well for many companies, but the odds of success are bound to increase if IT gives its blessing -- and support, when needed -- to software-as-a-service solutions.
Another reason to put business considerations first, rather than technical ones: Most commercially available tools satisfy basic technical requirements such as allowing integration of data from multiple disparate sources or providing the ability to specify organizational hierarchies. Those types of IT-driven requirements miss the broader business point because they don't address the specific types of business information, analytical techniques or decision support that will be used or even the core business processes a company hopes to improve with BI.