Coach’s Corner – Andrew McFadyen, Executive Coach
Pulling Out the ROOT of All Evil
In Lean terms, we all know that Waste = Evil. In all of its forms, Waste takes you and your business down. It makes you inefficient, frustrated, and fights against your fundamental purpose. Waste = Weakness. Just about every lean practitioner can describe many instances of waste that have taken hold in their business processes, production operations, or service delivery. As adept as you have become at doing kaizen and finding solutions to drive out waste, I notice that one cornerstone Lean tool is generally not practiced strongly. That tool is Root Cause Analysis – the Fishbone diagram.
Greenbelts and Blackbelts have used the venerated fishbone many times throughout their training, and hopefully afterwards as well. They can explain why it is important to dig deep and understand the root cause of problems so that the symptoms do not come back as you turn your attention to another problem to solve. However, we coaches notice that real skills in using root cause analysis only comes after dozens and dozens of uses, when people open up to new ways of thinking and patiently consider many alternatives.
Have you noticed that young kids ask “why” frequently? We may find this incessant curiosity is cute or annoying. But actually it is crucial to a child’s learning process. “Why” is one of the most powerful learning words. When a child asks “why?”, they realize they don’t know something and are eager and open to learn about it. Any parent knows that kids are REALLY EAGER AND OPEN TO LEARNING! Their minds are not pre-disposed with hunches, and they don’t feel so clever that they should already know the answer. But when adults ask “Why?”, there is often some presumption behind the question, or perhaps they are just interested in finding the first viable solution that comes along so they can feel good about notching another accomplishment on their belt and get back to their busy day.
I remember once dealing with a recurring quality problem that was creating a lot of rework and hurting customer satisfaction. After a few days of up and down results, I gathered the stakeholders together after the urgent guess & check / quick fix approach was not working. We pulled a large white board out into the Gemba, and I drew up a BIG fishbone. First, we discussed the problem and wrote a clear problem statement in the head of the fish. Next, we started populating the body of the fish, going bone by bone to consider “could-be” alternatives. This is the Lateral Thinking part of root cause analysis—to withhold judgement and develop a broad range of possible streams that fall into each fishbone category. Indeed the bones of the fish exist to make us think of different types of causes, so that we don’t get stuck on the first idea that comes up.
Of course a few people wanted to cross some items off the fishbone right away since they had experience or had done a little checking, but upon deep questioning, we realized they did not have sufficient data to rule out a single one! This was important – to challenge our belief that it must be one thing or it couldn’t be another thing. We should only ignore a potential cause if excluding it is supported by data and conclusive experiments or investigations.
Now with a broad set of possibilities up and visible, we then prioritized which items we thought were most likely to influence the quality problem. This is the first time we allow judgement to come into the process – not to rule items out, but to decide which items to check out first. We selected three possibilities to explore deeper: two Method-related items and one Material-related item.
With the short-list decided, we next started asking the 5-Why’s (i.e. why would that happen?). I realized that the team’s input was very important at this point. As a facilitator, it was crucial to recognize when a comment implied that a branch was possible, or when to stay on a linear track. Example: the temperature might have gone too low. Why? The oven controls allow too broad a range of variation, OR The operator did not check and allow the incoming material to warm up to room temperature before processing it = Two different branches off of the temperature stream. Showing them both does justice to them. Dismissing one leaves us open to tunnel vision.
Why Five Why’s? That is a good question. Of course it is a rule of thumb. We teach Greenbelts and Blackbelts to keep asking “why” until they get to a “Systematic Problem”. When the answers switch from “Symptomatic” responses to “Systematic” under-pinnings, then we have peeled the onion back and are starting to see the core. So the art of asking the 5-Why’s is to state the Why Question in a way that leaves us open to discovering systematic problems. Example: instead of asking “Why did we let the maintenance guy go on vacation when the filters might need to be changed?”, we should ask “Why didn’t we have a way to get the filters changed regardless if the maintenance guy was here or not?”. Asking good 5-Why questions takes practice. So practice it! Think of examples in your business where symptomatic problems can lead to systematic improvements by asking “why” the right way. So why 5? Well, frankly, it is just a rough average. Sometimes it takes 2 Why’s. Sometimes 10 why’s. Here’s how to gage if you are being silly with your number of why’s. (1) If your answers continue to look symptomatic after five why’s, consider what system is broken, and who ultimately controls it, then incorporate those aspects into you next answer. Example: Hmm, HR is responsible for the vacation system, but actually the production supervisor is responsible for manpower assignments, so maybe the filters didn’t get changed because we don’t have a system to ensure skill coverage when they need to be changed. Now we are getting somewhere. (2) Likewise, you can ask too many Why’s. If your answer sounds something like “because god created the universe”, then chances are you need to dial back a few why’s to see the root cause.
Advanced use of the 5 Why’s recognizes that you don’t always conquer Rome in a day. Sometimes it takes two days, or a week, or a year-long siege. When running the 5-why’s, experienced practitioners know when to recognize the need for more information. This is a critical point that requires some patience to say “we don’t know; let’s go and get some data”. The insight here is that sometimes you have to stop a 5-why stream and check with a subject matter expert or run a scientific experiment to prove/disprove the hypothesis.
Ok, back to our story. After a particular test on processing method, the results were inconclusive. Time to give up? No! We figured out what the noise was and ran another experiment. Then we could see a weak correlation with the quality problem. Not enough to call it a root cause, but still a source of variation that should be controlled. At this point two weeks have passed and we still have the Fishbone diagram open, and are reviewing progress with it daily. It has become a launching pad for daily investigations and activities. We are determined to solve this problem!
Finding a similar results weak correlation with another Method stream, we then turn our attention to the material stream. The Production Manager proclaims “No-Way! It can’t be the material. We are operating to all the material standards and all the checks are meeting specifications. It’s got to be the downstream processes”. Quite a strong statement from an experienced Manager would put some teams off and cause them to look elsewhere. Fortunately though, a good root cause process and Kaizen culture do not accept that. Indeed, a junior smarty-pants engineer had the courage to ask “why do we know it isn’t the material? How do we measure that?” The engineer was on to something. He knew the Production Manager’s math didn’t quite add up. Just because we were meeting prescribed specifications didn’t mean we were checking the right things. So he challenged Production to measure some different non-traditional parameters that might reveal an important source of variation in material composition. Grudgingly, the Production Manager agreed, but not without muttering “waste of time, won’t find anything”. He was relying too much on his experience and not using his sense of curiosity. So the engineer arranged to take measurements and sort the material into stratified batches, and then process those batches in groups. It took another week, but the results were dramatic and conclusive – one batch with a particular material profile was virtually defect free, while all the other batches were riddled with defects. We repeated the experiment and confirmed the result. We had found the root cause in material composition! The Production Manger even reflected how shocked he was, and how impressed he was that the young up-start had challenged his conventional thinking. Now all we had to do is control it, which we did, and quality shot up.
So what have we learned from this story about Root Cause Analysis?
- PRACTICE IT! It takes time and practice to use root cause analysis well
- DEFINE THE PROBLEM CLEARLY. Use quantitative metrics where possible that show a real gap
- GET IT UP AND VISUAL. Root Cause Analysis is a natural TEAM exercise
- WITHHOLD JUDGEMENT. Think laterally and brainstorm a broad set of possible streams on the Fishbone before concluding anything
- ONLY EXCLUDE ITEMS BASED ON DATA OR CONCLUSIVE EVIDENCE
- BE PATIENT. It may take weeks to check out the various fishbone possibilities
- BE SCIENTIFIC. Run controlled experiments to isolate potential root causes and see their influence. It may take multiple iterations to get conclusive data
- CHALLENGE ASSUMPTIONS. Do not just rely on someone’s summary experience. Check it out
- BE LIKE A KID. Ask lots of Why’s with the intention to learn something new
- ASK GOOD WHY’s. Probe for systematic problems
- RECOGNIZE BRANCHES. When a stream of inquiry can go two ways or more, show all those ways as separate streams and see where they take you
- KEEP IT ALIVE! Stay at it until you squeeze out the truth. Then find the control point and put in place solid sustainment standards
So there you have it. A dozen ways to help you get to the real root cause.
Return to Lean Sensei’s Newsletter Q4 2015 Issue