“Revered for decades as one of the world’s most innovative companies, 3M lost its innovative mojo when it began using Six Sigma to try to improve its operational efficiency. James McNerney, the CEO named in 2000, was a Jack Welch protégé from GE. He introduced the Six Sigma discipline as soon as he took the helm of the firm, streamlining work processes, eliminating 10% of the workforce, and earning praise (initially) from Wall Street, as operating margins grew from 17% in 2001 to 23% by 2005.
But when McNerney tried to apply the Six Sigma discipline to 3M’s research and development processes it led to a dramatic fall-off in the number of innovative products developed by the company during those years.”
Sourced through Scoop.it from: www.linkedin.com
Michel Baudin‘s comments:
Don Peppers describes “eliminating 10% of the work force” as part of implementing the “Six Sigma discipline,” but I don’t recall seeing anything on that subject when learning about Six Sigma.
It sounds much more like the rank-and-yank approach to human resources, in which every manager, every year, is forced to grade his or her subordinates on a curve and fire the bottom 10%. Jack Welch introduced it at GE, and it was propagated by his disciples at Microsoft, eBay, apparently 3M, and many other companies.
The original Six Sigma was nothing of the kind. It was a modernization of the old, 1920s vintage Statistical Process Control (SPC), using statistical design of experiments to help manufacturers make their processes capable. And part of the approach involved training a cadre of “Black Belts” in a set of tools.
I have no doubt that it fulfilled this purpose but then it was taken global-cosmic, presented as a general approach to management, and commingled with rank-and-yank.
That, as such, it should do more harm than good in R&D is no surprise, simply because this activity thrives on collaboration, which is discouraged where helping a peer succeed may get you fired. But even the original Six Sigma is irrelevant to R&D. Inventing is not about making a process stable.