Statistical Method Helps Boost Bottom Lines, Batting Averages [the broken link was removed] by Jon Van, Chicago Tribune:
Using statistical techniques embraced by quality guru W. Edwards Deming, Holland has worked for a generation guiding enterprises large and small to boost efficiency.
…
Yet getting management support for an MVT project is vital, Holland said, because experience shows that only 25 percent of ideas intended to improve a process will have a positive effect. The others will either have no effect or will hurt. Managers hate to see experimental results shoot down ideas they were certain would help, he said.
The article also mentions:
While Holland claims MVT as his own, others say it is really just a variation of strategies widely used in business.
“Multifactor experiments have been around for a long time,” said Ajit Tamhane, Northwestern University professor of statistics and industrial engineering and management sciences.
…
David Coit, a Rutgers University professor of industrial systems engineering, said that Holland’s MVT is very much like a quality-enhancing scheme called design of experiments.
So often we seem to focus on proprietary solutions. Instead it seems to me, most often what is needed is to do a good job of applying the ideas that have been known for decades. Deming ideas, design of experiments, lean thinking, experimentation, etc. are not secrets. There is a long history of how to apply these ideas to improve organizational performance.
QualPro obviously does well marketing itself (see press clippings from their web site) selling the concept of proprietary solutions to press organizations. Raising the question of whether the proprietary solutions really offers unique ideas, as the The Tribune article did was uncommon in my experience.
I like those encouraging the adoption of statistical tools to improve management but I find the practice of trademarking terms like Six Sigma and MVT a bad way to encourage innovation in the practice of management. While it is nice to see Six Sigma efforts and others use statistical tools (such as design of experiments) I would encourage people to stay with “open source” management terms and remain part of a community looking to improve the practice of management.
Update: Also see – Management Advice Failures
