Finding the Root Cause of What's Right
W. Edward Deming said the most important figures in any business are invisible, unknown, and unknowable. Peter Drucker added, These results exist only on the outside, such as an extremely satisfied customer. Similarly, the only thing you can measure on the inside are costs.
Both Deming and Drucker use the "invisible figures" idea to show that managing is a liberal art, employing a number of strategies simultaneously to help aim toward success. This falls in-line with the Shewarts Cycle of Continual Improvement: Plan-Do-Study-Act. We are ever looking for better process measures to tell us what to improve.
While improving our processes, the question arises, How do I know what I don't know? Being able to see the blind spots of our business is the central challenge of managers, engineers, analysts and knowledge workers. The following model helps to illustrate the problem:
The question jumping out of the picture above is, with so little knowledge of all possible factors in the universe, how do I know what to fix? Put another way, How do I know what I don't know -but need to know.
There's an old tool many of us are familiar with called Root Cause Analysis, or the 5-whys. For a simplified description of RCA please see Hassan Zahran's article here. We are very good at employing this tool when something goes wrong, but what about when something goes right? If you are only using root cause analysis when things go wrong, then you are only getting 50% of the benefits.
Here's an example of Root Cause Analysis on things that go right:
The results jump outside the previous process limits for each of the circled improvements above. Each time that happens, the data is telling us that we have a new root or assignable cause whether we were previously aware of it or not. This is the time to start investigating on what's going right in the process. For assignable causes, the answers may be obtained by interviewing the production worker. These lessons must be captured and standardized in order to remain durable. If not, the improvements will likely go away when the operator moves or equipment is replaced.
If you are using root cause analysis only when things go wrong, then you are only getting 50% of the benefit.
The next level is to start digging into common cause events. It's been said that common causes are assignable causes that haven't been discovered yet. By taking the time to investigate the improvements in our process, we can make those improvements durable and continuous over time.
Good luck in applying root cause analysis tools to what goes right in your business!
Volunteer Guest Lecturer at US Navy ROTC, Retired from the Aviation Connectivity and Communications Industry
8 年Well said. It is out of balance. What goes right is ignored and we spend far too much time analyzing the failures.
Mission Assurance Lead at Advanced Thermal Batteries
8 年We should all be so lucky to work in an organization that is appropriately staffed and managed (and encouraged) to take positive RCA action. Unfortunately - in many cases the "overhead" of quality in many organizations that are struggling in the new economy, barely alot the time and resources to adequately address the problem RCA much less the upside. I am, however, forwarding this excellent article to my boss. Thanks Joel!
Chief Operations Officer QP & CQ Manufacturing
8 年Found myself asking that question today. "Hey, it was a good day today! Why? We should do that all the time! Ok...so how do we make that happen?" With data, sure it means that ability to see mean shifts, and to identify trends, but sometimes it can be broader than that. I'm in the situation of uprooting a quality system and replacing it with a leaner, more robust system, and found myself asking "what causes a quality system to be effective?" Not suprisingly, I went to a fishbone diagram to account for the factors that go into a successful system, and for each of the major factors, I cause mapped their contibutors. Sometimes employing an old familiar tool, ordinarily thought to be engineering oriented to the management system can cause you to look at things in a new light. Certainly helped me gain some clarity on what tasks needed accomplished next in invigorating our system.
Corporate Quality Sr. Engineer | Hitachi Astemo | MS in Engineering Management
8 年Traditionally RCA is implemented only when things are going wrong but as this article compels I believe RCA should also be implemented to understand why things are "not" going wrong. This would help us to identify factors that would help 'standardize' a process. Achieving Standardization in a process achieves Quality. Great insights about Quality Engineering in the article.