When you discover a problem in your organization, do you employ root cause analysis technique such as Five Whys or Fishbone? You should. If not you risk applying superficial solutions that treat symptoms, rather than eliminating the underlying drivers of problems.
Numerous online examples will demonstrate the basic techniques. Applying them to issues where you have first-hand knowledge is easy enough and will make you familiar with the approach. Root cause analysis is easy to understand. But performing good analysis is more difficult. It requires both practice and a way to gauge the analysis quality.
Here are two tips for testing and improving the quality of your root cause analysis. These tips are both simple and powerful. Apply them consistently and you will soon wonder how you did without them.
Tip #1: Scatter, Then Gather
When looking for causes at any level in your analysis, generate ideas first, without filtering them. Come up with several alternatives. Maybe each is a contributing factor. Don’t gauge the strength of the contribution immediately. Just note them all down. Only after generating ideas, move on to determinng relative strength of the contributions, and drilling down on the strongest ones.
Separating the generations of ideas from the filtering of them (in this case determining strength of contribution) is a scatter-gather pattern that usefully applies to most activities with a brainstorming component. It keeps you from closing down options prematurely, and gives you a broader set of contributing causes to consider. It ensures proper breadth of analysis.
Tip #2: Read Down, Then Up
Once you have assembled your chain of causes, read it aloud from top to bottom, from your initial observed problem down to the root cause, joined on “because”. This happened, because cause number 1, because cause number 2, … Such reading may reveal wrongly ordered items, items that are not truly causes, causes that are weaker than first believed, and so on.
Then read aloud your chain of causes bottom to top, from the root cause up to the initial observed problem, joined on “therefore”. This cause happened, therefore consequence number 1 happened, therefore consequence number 2 happened… As with reading down, this may reveal flaws in the analysis, weakness in the causes, or some other shortcoming in analysis.
As simplistic as this technique sounds, it will provide you immediate insight into the quality of your inferences. When weak inferences arise, redo the analysis. Iterate until your chain reads well both down and up.
Apply these techniques consistently, every time you do root cause analysis. The quality of your analysis will improve quickly, and you’ll be on to the next problem.
Posted by William Baxter on Jan 17, 2016