Those of you who regularly read my column probably realize that I sometimes like to take an unpopular position in
order to stimulate readers' thoughts and imaginations. There's nothing more boring than reading an article that makes you say to yourself: "Oh, yes, I agree; I've said that a number of
times. Yes, that's the way it is." We all need to get out of our boxes once in a while and challenge our own thinking. In the June issue of Quality Digest, I tried to
stimulate some reader response by questioning the value of Six Sigma. I offered a free performance improvement book and CD-ROM set, to the first five people who provided me with a
publishable case study of a service process that was performing at the Six Sigma level. I thought I'd be deluged with hundreds of excellent examples that would be well worth my investment. Well,
I still have four of the book and CD-ROM sets. But--surprise, surprise--I was deluged with e-mails agreeing with my negative position and only one "hate mail" from a Black Belt
consultant, who was angry because I dared to question the Six Sigma approach. Because readers weren't able to provide me with examples to share--although I'm sure there are
good ones out there some place--my planned theme for this month's column, "Six Sigma in Service," had to be changed. Instead, I'm writing about the evolution of Six Sigma as a
measurement standard. In 1924 Walter A. Shewhart defined "three sigma" as the point in a process that requires correction. The next development in acceptable quality
performance was the Dodge-Romig sampling plan, which converted statistics into a simple table. During the 1940s, a Cpk of 1.4 became an acceptable performance level to qualify a piece of hardware
or test equipment. A complete turnaround in acceptable quality performance, known as Zero Defects, was developed by Philip Crosby in 1961. This was a major breakthrough
because, for the first time ever, no level of defects was acceptable. Essentially, Zero Defects was a 12-sigma approach because, statistically speaking, at 12 sigma, the population required
before a predictable defect could occur is so great that almost none would. Phil's concept was quickly embraced by the U.S. government, which sought defect-free products from all of its
suppliers. Zero Defects fast became the "in" thing to do and took over the United States like the Beatles did during the same era. This concept marked the start of the continuous
improvement requirement. As we entered the 1980s, service became a major part of the U.S. gross national product. To apply Zero Defects' improvement standard to service and
support areas, the phrase "Zero Defects" gave way to "error free." Thus, another breakthrough occurred as tools such as statistical analysis were applied to research and
development, personnel, sales, schools, government and others. This spurred on the TQM movement, which was a "womb to tomb" approach to massive improvement, largely driven by the superb
quality built into Japanese cars, steel and ships. In the early 1980s, Motorola sold its Quasar TV manufacturing and engineering facilities to Matsushita and was amazed when
the quality level improved dramatically. Motorola realized that it, too, needed a massive improvement in quality. The company's former CEO, Robert Galvin, and his team set a target of 10-times
improvement in five years. Of course, this established a performance target that was less than the then-present standard of zero errors. The Six Sigma standard was established to support Galvin's
directive because few people believed that perfection was possible. The Six Sigma process applied well to commercial products for which a certain level of errors was acceptable and reliability
was not crucial. With the Six Sigma measurement standard, the reality of human performance drew more in line with a measurement standard. For example, I bought a set of 44
five- to 15-minute videos about Six Sigma put out by the American Society for Quality and developed by the Six Sigma Institute, and they had many errors: tapes that wouldn't run and directions to
use items not included. Apparently this was acceptable performance under the Six Sigma criteria because when I reported the errors to ASQ and the authors, only the tapes that wouldn't play were
replaced. The nice thing about Six Sigma is that it's a goal that might someday be reached, but you don't have to be there with the output you're delivering today. Today's
output might not even meet the three-sigma requirement that Shewhart designed into his control charts during the 1920s. The Six Sigma tapes are a good example. They fell far short of meeting my
needs, let alone customer expectations, even though they cost thousands of dollars. To err is human, to be paid for it divine. To repeat the same error is insulting, both to
management and customers. About the author H. James Harrington is COO of
Systemcorp, an Internet-software development company. He was formerly a principal at Ernst & Young, where he served as an international quality adviser. He has more than 45 years of
experience as a quality professional and is the author of 20 books. E-mail him at jharrington@qualitydigest.com . Visit his Web site at
www.hjharrington.com . |
