Problem Solving Techniques and Strategies I've Learned Over the Years

During my career I've had the opportunity to be involved in many problem solving situations that have forced me to "think out of the box." These situations have allowed me to increase my tacit knowledge to the point where I can now look back and see that certain patterns often presented themselves that gave me clues and potential starting points to help in my investigation. I'd like to shared some of these with you.

• First and foremost, keep an open mind. It's easy to jump to a hypothesis based on someone's opinion that can then start you down the wrong path. Don't! I call this group think. It's when someone throws out an opinion and everyone on the team jumps on board because they can't come up with a better idea. I had an experience several years ago when a team got so mired in things that they actually began to think there was an issue with how the molecules combined to form the compound they produced! They had been making this compound for several years and now all of a sudden there was an issue with how the molecules combine! It turned out the problem was caused by excess moisture in one of the ingredients in the compound.
• Don't take anything for granted. Make sure you can validate every assumption and hypothesis with data. Follow up on every lead. Try to make sense of every clue. Do all the clues taken together make sense? Do they point you in a certain direction? One of the toughest problems I encountered had a very low frequency of occurrence. A friend called me one day and told me she had one defective part and wanted to know if I wanted it or should she throw it away. I went to her desk to retrieve it and struck pay dirt. The part had "blue paint" on it and it gave me the clue that eventually helped me solve the problem.
• Go to the gemba and see for yourself. I recently coached a lean six sigma project at a hospital that involved a patient therapy issue that everyone was convinced was due to a patient scheduling problem. The two students leading the project went to the gemba and found that it wasn't a scheduling problem, but a problem with patients cancelling their therapy sessions and consequently required a different analysis and solution.
• Torture your data to give you clues. Sometimes you're blessed to have initial data. If that's the case (and you can trust it!) make sure you get as much information from it as you can. An example of this occurred many years ago when a team convinced upper management that they were having a product damage problem due to shipping. The team spent a considerable amount of money improving the shipping container and product packaging, but the problem persisted. I was able to get some of the original customer complaints and doing some simple pareto analysis determined that over 40% of the defects were coming from two operators on second shift. It was a no-brainer! After watching one of the operators from a distance for 10 minutes, the supervisor and I observed him throwing the product into the container and were able to validate this as the root cause of the damage.
• Sometimes you can back into the problem. If the problem occurs frequently, you have a good chance of seeing the conditions that create the issue in real time. You can use some simple tools such as concentration diagrams, 6M analysis, and check sheets to help in your analysis and investigation. An example of this occurred when I was asked to go to a Mexican plant to investigate a product failure problem. As we watched the operator assemble the product we were able to see the failure mode in real time. Doing a comparison of good product vs. bad product, we determined that when the operator used the correct method good product resulted. If they altered the method, bad product resulted. And if the problem occurs infrequently, you still have the opportunity to document the conditions which contribute to a good outcome while you wait for the variation cause to effect the process. A useful tool to use in this case for variable data is a multi-vari chart.
• Use a strategy of shift and squeeze when you're trying to reduce overall variation or shift the mean. You can use tools such as fishbone diagrams and 6M analysis to determine which of the M's is affecting the process. Then it's a matter of analyzing the causes and identifying potential solutions to reduce the variation.
• I always start my problem investigation using 6M analysis. I go through the mental checklist looking at each M: Man, Machine, Material, Method, Measurement, and Mother Nature. This strategy has served me well over the years. Several years ago, a client had a large accounts receivable problem in their European operation. They had assigned the problem to a new Black Belt. She and her team were struggling and not making much headway. As I went through her analysis, I kept asking how they escalated a receivables issue? How big did it need to get before it was escalated to the next management level? What was the method and process they used? It turned out that there was no method or process and each accountant was on their own as to how they dealt with this customer issue.

Problem solving can be challenging. It forces one to really think about what's going on and how different variables and situations can effect the process. Being patient, asking questions, and keeping your eyes and ears open to how things operate are some important traits to possess.