Friday, March 20, 2009

"Making Six Sigma Faster"

I recently had a manager tell me that his company had decided that Six Sigma just “takes too long,” and that they were implementing a new mandate to “Make Six Sigma Faster.” I asked him why they thought Six Sigma takes too long. He told me that the average time to complete a project there had been 8-9 months; under the new program, Black Belts were going to be expected to complete projects (at least through the IMPROVE phase) in 90 days or less.
I couldn’t help it…I broke out laughing, and asked him “By what method?”
“Well,” he said, “we’re not sure yet, but we think if they start holding more meetings, and never go into a meeting without having the deliverable already drafted, that will help.”
I had asked Deming’s famous question for a reason. I was very familiar with this particular deployment, and I knew that the reasons their projects always ran long were many, but almost none had to do with the Black Belts or the number of meetings they held.
This organization had decided at the beginning that they just didn’t have time to do a couple of days of Champion training. They had decided instead that they could get along with a 2-hour teleconference and a required reading list. This was a big organization, and they had never bothered to set up any kind of listening posts or other pipeline-feeders, had no project portfolio management, had not coordinated with the PMO, hadn’t trained any middle managers in SPC, Six Sigma familiarization, or anything else. Black Belts were pretty much expected to find their own projects, and in many cases had to hunt down anyone willing to sign on as a Champion (sometimes, they just picked another Black Belt, because “at least I got someone who understands Six Sigma.”) Getting good data was another problem. Usually, there were no data available for even deriving a baseline, much less for stratification or for digging into cause systems to find “x’s.” Just collecting the data for a baseline might involve a couple of months’ worth of work. The organization often relied for stratification on “reason codes,” the use of which were consistently proven unreliable when tested using attribute agreement analysis.
These were the primary factors driving project lead times, but there was no plan to deal with these factors, because it meant getting leadership to change, and no one at my manager friend’s level had the ability to push that noodle uphill. So they were just going to do the usual thing…put it into the expectations for the Black Belts. All you have to do to get the variation narrower is tighten the specs, right?
This was Deming’s point; to paraphrase, “If you could cut the project time by 60 percent this year without a method, then why didn’t you do it last year? Must have been goofing off…”
Listen, executives: This is too important. Six Sigma, implemented properly and led from a systems perspective, is a proven methodology that will make your business better, your customers happier, your revenues higher, and your costs lower. But it’s not something you bolt on, walk away from, and just wait for the cash to roll in. You have to lead it, you have to be engaged, you have to remove obstacles and make Six Sigma a strategic component of a larger Quality Management System. In the end, if it fails, you can’t blame the Black Belts you didn’t support, or the culture you didn’t change, or even the consultants you didn’t listen to. It’s not that it won’t work at your company…but it certainly won’t work if you don’t lead it!

No comments:

Post a Comment