Thursday, August 6, 2009

Back Again

Sorry I've been out of this loop for so wife is fighting a life-threatening cancer, and dealing with that just sucks every ounce of spare energy and time out of your life.
I'm going to jump back in by recycling my latest answer in the Deming discussion group on LinkedIn. The question was about Six Sigma and Deming. Two of my favorite Deming Disciples, John Dowd and John Constantine, feel that Six Sigma is fundamentally flawed and has very little place in any discussion of Deming. This is what I wrote:

I generally agree with John Dowd and John Constantine on most things, and I agree that Six Sigma-- as taught by many of the companies consulting in it these days--has some serious problems. Some of these approaches are, indeed, incompatible with the Deming philosophy. And some of the consultants using those incompatible approaches are large enough that the argument for "as generally taught" is probably sound. My suggestion, though, is that it doesn't have to be that way. Six Sigma is just a marketing vehicle; there is no standard for it (although ASQ would like to think that their SSBOK is one). As with TQM and every other quality approach there are people who do it well, and people who don't. Unfortunately, those who don't could care less about transformation, because many of them know very little about variation, and next to nothing about systems theory, psychology, or theory of knowledge. My disagreement is in what we do about this. I have been doing what I could, through my consulting practice, any conference appearances or workshops that I am able to do, and any writing that I can get published, to bring Deming principles into Six Sigma, and to use an approach in Six Sigma that is consistent with Deming. Clients who work with me only calculate the "1.5-sigma shift" and the "process sigma" as a curiosity and a metric for communicating with those less enlightened. They see and discuss the Red Bead, the Funnel, systems theory, SoPK, and the 14 points. They learn to see and use Six Sigma projects as one tactic in an overall process management system. I think if more Deming practitioners could find it in their hearts to do something like this, we'd have more success in using Six Sigma as one tactic in our overall aim, and reach more managers and executives (and potential managers and executives) with our message. This is what Lou Schultz and William Scherkenbach taught me many years ago. As to "picking a target and pretending to improve the system by keeping centered on it," I think we need more context. It's true that arbitrary targets are harmful, but this is really the basis for world-class quality; Taguchi defined it in 1960 as "on-target with minimum variation." Getting a process centered on its specified nominal value and constantly reducing variation around it has long been the the goal of anyone trying to understand variation and use that understanding for improvement. I don't think that has changed; used properly, DMAIC projects can help a team achieve the kind of fundamental changes to a system that are needed when the process is stable but off-target or out-of-specification.