BI and Analytics Centers of Excellence: Current Models No Longer Making the Grade

ArunRangamaniBy Arun Rangamani, SVP, Analytics & Technology, SCIO Health Analytics
Twitter: @SCIOanalytics

A business intelligence and analytics center of excellence (CoE) is an organizing mechanism to align people, processes, technology and culture. CoEs are intended to bring about better collaboration between business units and IT; increase adoption and use of business intelligence and analytics; enhance data management, quality, and reporting; and save costs by eliminating redundant functions.

The problem: Current business intelligence and analytics CoEs are not quite what they once were. In fact, as the healthcare environment continues to change – and a variety of constituents demand more knowledge – these CoEs are well past their glory days.

In fact, common business intelligence and analytics CoE models being used by analytics vendors are typically product based “uni” solutions, while organizational IT departments typically address business intelligence through a group of homogeneous staff members. While these models worked in days past, they are no longer scoring the winning touchdowns in today’s healthcare game.

Consider the following scenario: A hospital chain is looking for insights on both the causes and costs for readmissions and what can be done to prevent them. Under established, prevailing CoE models, a business intelligence product vendor would read data marts and queries that were provided by an analyst – and based on inputs from a group of homogenous analysts at the hospital. Alternatively, the hospital IT staff would try to address the issue through a business intelligence solution, but would naturally struggle with bandwidth issues.

Such models simply don’t work because:

#1: They plod along too slowly. Current CoE models are not capable of producing impactful business intelligence insights to constituents in a timely manner. Indeed, these models cannot come up with the desired analyses in an efficient manner because the analysts working under these systems do not bring the appropriate mix of skills required to get the job done efficiently and effectively.

#2: They can’t handle what’s thrown at them. These models are not capable of managing multiple data sets. However, big data is converging upon healthcare. Therefore, healthcare organizations are dealing with multiple sets of complicated data. The lack of a proper plan amongst the execution/delivery teams regarding what questions to ask the data makes it difficult to produce desired results. In addition, the lack of the right skillset makes it difficult to build the right business intelligence delivery framework.

#3: The output doesn’t win the game. With these models, outcomes most likely just provide descriptive reports with very little insight. As such, the growing need – prompted by government requirements and other industry forces – to obtain more detailed reporting with insights are not met.

While these models have definitely seen better days – the question is how exactly can new, emerging business intelligence CoEs meet today’s challenges? It’s time for us all to think about that. In fact, I will pick up the discussion and offer my thoughts on the topic in my next blog. Until then . . .

This article was originally published on SCIO Health Analytics and is republished here with permission.