Getting Results with the Fishbone Diagram

The Fishbone Diagram is easily one of the most applicable tools in your Lean toolbox. It really helps to grasp all of the potential root causes of an issue that you’re seeing in your manufacturing operation. But there are some limitations. Notice I used the word “potential” root causes.

The Fishbone Diagram is often misused as people believe that it reveals the true root causes of issues after one pass with the tool. However, this is a misconception because the tool is simply a method of facilitating a broader learning and discovery process. The first pass is no more than a brainstorming session for all the factors that could be related to the problem. It’s what happens next that determines if you actually ever identify the true root cause.

The drawback of the Fishbone Diagram is that it doesn’t intuitively lead users (especially those who are inexperienced) to apply the scientific method. Most people intuitively believe that if they fix all the items identified on the diagram, then the problem will be solved. The reality is that every item on the initial cut of the diagram is no more than a hypothesis; each needing to be tested, evaluated for results, and experimented with to see exactly how much it actually affects the visible problem.

The scientific method includes:

1. Forming a hypothesis
2. Collecting data to evaluate the baseline condition
3. Making changes in a controlled way
4. Collecting more data to prove or disprove the hypothesis
5. Drawing a conclusion based on the findings

The Fishbone Diagram is an excellent way to complete step 1 of the scientific method. In fact, it can even provide you with a dozen or more hypothesis, all of which to be tested. This could be expensive and even impractical. So how do you get real results using the Fishbone Diagram in an environment where you have finite time and resources to commit to a given problem?

Well, the answer is in prioritization. First, pick a really important problem – preferably something that’s tied to your company’s strategy. For example, if maximizing quality gets you ahead in the market or allows for increased profit margins – or whatever else is most important for your business, then pick the problem that has the greatest negative impact on product quality. A pareto analysis tool can be used to help in this case where issues are quantified and ranked from greatest to least.

Second, execute the initial fishbone discussion with a cross-functional team of people close to the issue. This may reveal quite a few hypothesis, which all need to be validated. From here choose the most likely few items that you’d like to test. You can have the team vote or have a quick alignment discussion to gauge what people think are the most likely root causes. Then attack the items that get the strongest response. If they don’t produce the result you’re looking for, go back to the diagram and call out the next few most likely root causes to be worked. The real secret sauce is to not give up until you have unquestionably solved the problem you set out to solve.

Part of your testing should apply the logic of 5 Why Analysis. Some issues are caused by a chain of events and the 5 Why’s is a powerful tool for tracing issues back to their true root.

Combining the Fishbone and 5-Why approaches leads you to my favorite approach to Root Cause Analysis, called the Why-Tree. It gives you the best of both worlds between the Fishbone and 5-Why’s. The Why Tree reveals multiple potential root causes like the Fishbone and applies the linear logic of the 5 Why’s. In a brief 45-minute discovery session with a cross-functional team, the Why Tree can reveal incredible insight about manufacturing or other business problems.