Over the last year, this column has looked at several interesting business intelligence (BI) trends that highlight the tremendous potential of the data warehouse. Operational BI (embedding BI into the operations of the organization), event-based marketing (triggering customer communications from behavior pattern changes) and contact optimization (ensuring timely, relevant communications) are a few of the powerful applications that we have touched on. However, I have not mentioned that along with success comes heightened visibility, and if we are lucky, increased demand for new capabilities.

One mechanism that data warehouse teams employ to stay ahead of the curve is the BI center of excellence (COE). As with other popular BI terms, there are many definitions for COE. I have two favorites that I think complement each other quite nicely. The first definition is one that my organization, Intelligent Solutions, Inc. (ISI), developed: “The COE is the set of people, processes and technologies for promoting collaboration and the application of best practices. This includes methodology, best practices, education and training, support services and technology awareness.”

I think of this as the working person’s definition of a COE. It reflects the nuts and bolts of what must get done to allow the data warehouse team to expand as needed to meet increasing demand. A primary role of the COE is to define and document best practices, templates, training programs, communication programs and standards. Unfortunately, this takes time - particularly the defining part. The more people that are involved, the more potential there is for differing standards, lack of documentation, conflicting definitions and overlapping (or worse, underlapping) processes. This is particularly true when the list of projects is long, resources are scarce and potential for becoming a bottleneck is high.

Thus, as our definition highlights, it is imperative to establish the discipline for processes and documentation early to ensure that new innovations, tools and technologies follow existing standards. This will allow the COE to expand mentoring activities, reduce training times and allow resources to become productive faster. Processes and documents that should be established up-front (and refined as necessary) include data acquisition and metadata, business requirement gathering templates, change management processes and training programs. Additionally, skilled data warehouse resources are in short supply in today’s market, and high demand for COE resources along with constraints on internal resources may cause the COE to become a bottleneck. A popular workaround for this common problem is to contract external resources in times of high demand. The methodology and best practice documentation that we call for is an effective and efficient way to ensure external resources are performing to the same standards as the rest of the COE.

A complementary definition for a COE is one that I picked from up from conversations with a vendor in the BI technology space, SAS Institute, Inc. Where ISI’s definition details the nuts and bolts, SAS’s is a bit higher level, focusing on several key concepts I feel are also critical: “The COE is composed of formal cross-functional team(s) that have clearly defined roles, responsibilities, and processes and that drive BI through the enterprise.”1

This definition makes my favorite list for two reasons. First is the use of the phrase “formal cross-functional team(s).” The word formal implies that the COE is an actual organization with a defined structure. It also implies that the planning attendant with the development of a formal organization unit - including mission, charter, roles and responsibilities - has been done (more about this in a future column). Further, formal is a way to ensure that the very real duties of the COE, including the unglamorous but essential definitional work we discussed, are formally assigned as a high priority rather than simply being a task at the bottom of someone’s priority list.

The term cross-functional implies that more than one part of the organization is involved in the COE. This is key. An effective COE must represent both the key users of the data warehouse (business) as well as the key implementers of the data warehouse (technology). This only makes sense if you consider that most of the emerging uses for BI I have discussed in this column require a closing of the loop between analytics and operations. The “s” in the word teams is also very telling. This implies that the COE might very well be composed of multiple teams rather than one large, far reaching organization structure. In my experience, particularly in larger organizations, it is the case that while the COE might start out as a single team, it usually evolves to include multiple teams, each with a specific focus. We will take a closer look at some of the potential structures for the COE in a future column.

The second reason I like this definition is the concept that the COE has a role in driving BI through the enterprise. I couldn’t agree more. The COE is more than a group that builds the data warehouse environment. Instead, it must be structured such that it can ensure BI activities are in line with the strategic direction of the organization. Over the next several months, we will examine the COE in more detail, highlighting critical success factors, describing options and discussing common roles and responsibilities.


  1. Gloria J. Miller, Dagmar Brautigam and Stefanie V. Gerlach. Business Intelligence Competency Centers: A Team Approach to Maximizing Competitive Advantage. Hoboken, New Jersey: John Wiley & Sons, 2006.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access