Eckerson Group is hiring! We seek a Digital Events Coordinator and Podcast Producer. Apply here>

BI Power Struggle: A Strategy for Success

BI Power Struggle: A Strategy for Success

If you are a business intelligence (BI) leader and feeling on the defensive, there is good reason. Every ten years like clockwork, our industry swings from corporate-driven, governed data environments to business unit-driven, ungoverned data environments. We are now dead center in the ungoverned cycle. The good news is the pendulum is slowly swinging back and your services will soon be back in demand.

The figure below shows the two worlds of BI that compete for mindshare, money and resources in your organization. Rather than allowing the pendulum to swing from one world to the other, you need a strategy to plumb the middle and reap the benefits of both worlds without suffering their downsides.

Figure 1. Two Worlds of BI

Two Worlds of BI

Top down. The governed world is ultimately championed by the CEO who wants data consistency, the CFO who wants economies of scale, and the CIO who wants cost efficiencies. This world, which I call “top down BI”, is driven by the information technology (IT) department to meet the needs of casual users who primarily (but not exclusively) want to monitor known processes using dashboards that contain certified data gleaned from the data warehouse.  The downside of top-down BI is that generating certified data takes time, money and political willpower, and it’s not optimized to support self-service BI.

Bottom up. The ungoverned world is nearly the opposite of the governed world in every possible way. It is driven by business unit heads to meet the needs of power users who are tasked to find new insights (not monitor existing processes) and answer unanticipated questions using any data, certified or not. It is agile and supports self service to a fault. But the downsides ultimately infuriate the CXOs: lack of data consistency, cost efficiencies and economies of scale.


The top-down world dominated in the 1980s when mainframe-based reports and IT departments ruled the day. The 1990s shifted power to analysts in business units, giving them desktop analytical tools, such as Business Objects, Cognos, MicroStrategy and the ever-ubiquitous Excel. The 2000s saw the rise of dashboards and scorecards, graphical exception reports designed to help managers monitor key business processes. And today, we’ve moved into the era of prediction, idolizing data scientists and data engineers who find needles in data haystacks.

So, every ten years, the pendulum swings back. We’re now in 2015, and it’s interesting that we are now starting to hear the words “big data” and “data discovery” in the same breath as “governance” and “data semantics.” Yes, the tide is turning and that spells good news for much beleaguered IT and BI professionals who have been scrounging at the bottom of the budget bowl for the past five years.


Of course, waiting for your boomerang to return is one strategy. A better strategy is to figure out ways to knit these two worlds together into a coherent whole. They are really two sides of the same coin, and it’s a shame that in most organizations they end up competing with each other instead of complementing each other. You need both worlds, not one or the other, to effectively turn data into insights and action. That was the key theme of my last book, “Secrets of Analytical Leaders: Insights from Information Insiders.”

You can pull the two worlds together by rethinking your core architectures. At Eckerson Group, we help organizations assess, define and implement five architectures:

  • Business architecture. Do you run the BI program as a business? Do you treat users as customers and reports as products? Do you track sales and calculate market share? Do you have a business plan and board of directors?
  • Organizational architecture. Do you have the right people in the right roles with the right reporting structures? (Hint: the best organizational model is federated). Organizing for success is the biggest challenge.
  • Data architecture.  Do you flow data through your systems in an optimal manner? Do you have the right engines processing the right data to meet business needs?
  • Analytical architecture. Have you provided the right tools to the right people with the right support to handle tasks each needs to accomplish? Have you inventoried your user base to see who does what and when?
  • Governance architecture. Does your organization share a common vocabulary for describing the business and its performance? Does the business own and govern the BI program?

Mastering these five architectures will go along way towards managing the wild pendulum swings from top-down to bottom-up BI that can exhaust an organization and turn natural allies into enemies. The ultimate strategy for the new BI leader is to reconcile polar opposites and create a unified world that offers both governance and agility. 

Wayne Eckerson

Wayne Eckerson is an internationally recognized thought leader in the business intelligence and analytics field. He is a sought-after consultant and noted speaker who thinks critically, writes clearly and presents...

More About Wayne Eckerson