The Importance of Root Cause Analysis

On Tuesday, I chaired the North American Trading Architecture Summit. If you attended, I hope you walked away with new ideas and new contacts. In case you missed it, I along with my colleagues Tim Murray and Jake Thomases have written several analytical pieces from the event.

One presentation in particular caught my attention. Howard Halberstein, vice president and lead solutions architect of Deutsche Bank's UNIX division, provided a thoroughly entertaining case study on root cause analysis, in which he described two "causes" that relate to an event. Proximate cause is directly linked to the undesired outcome. He likened this to a missile hitting a plane, with the missile being the cause of the plane falling out of the air. Root cause is more difficult to determine as it is usually more political than technical.

Reactions to problems tend to be knee-jerk instead of the result of looking into the root cause. "Someone trips over a cord and we put armed guards in front of the datacenter. A person accidentally pastes something incorrectly, we remove ‘paste' from every computer in the entire company so it never, ever happens again," Halberstein says.

Determining root cause is difficult. "It's really hard in an IT shop to find root cause," he says, because a seemingly endless string of people, events and processes have to be examined. Numerous interactions take place, involving lots of people and different units and vendors—not to mention the data points to weed through.

I will explore this issue in more depth soon, but the idea that became clear at the event is that root cause analysis can only be effective with good quality data in place. And creating standards for where data is stored is essential. This is about creating efficiency throughout the organization and not just blindly throwing money or resources at an issue in a reactionary way.

I would like to thank all the delegates and sponsors who made the North American Trading Architecture Summit possible. It really was a great day.

If you have any feedback on the event, or want to discuss root cause analysis further, drop me a line at +1 646-490-3973 or anthony.malakian@incisivemedia.com.