At the 2009 National Association for Healthcare Quality conference, I gave a speech on lean Six Sigma simplified. At the end of the session, one of the attendees asked, "If Six Sigma is so easy, why isn’t everyone doing it?" My answer: Because we’ve made it too complicated, expensive, and hard to learn.
ADVERTISEMENT |
Seth Godin, author of Unleashing The Ideavirus (Hyperion, 2001, download it from www.ideavirus.com), says: "Ideas that spread, win." Six Sigma has been spreading, but slowly, at great expense and mainly in large companies. How can the quality community make Six Sigma easier to spread?
Let me begin with a story.
A woman I know went on a date with a man she’d just met. He took her on a picnic where they had to walk up a trail onto a small hill with a view of the foothills. I saw her recently and she was still dating the same man. She told me, “I don’t know how it happened, but six months later I was standing on top of an 800 foot pinnacle in Utah after a free climb.”
Her date was smart enough to know that he couldn’t take her free climbing on the first date. He had to gradually introduce her to increasingly more difficult climbs, technical gear, and so on. With each success, she built confidence that led her, six months later, to the top of a pinnacle.
Unfortunately, in Six Sigma, we seem to have forgotten that the average Joe or Susie has a deep seated fear of math, statistics, and technology. Instead of taking them on a picnic up a small hill, we throw them into weeks of Green Belt or Black Belt training. We spend days trying to teach them to use complex statistical software to do hypothesis testing and all manner of technical problem-solving tools.
Let’s face it; in search of big profits from training, consultants have made Six Sigma too complex, too time-consuming and too expensive for most businesses or people. Rather than lowering the bar to entry, we’ve raised the bar to ridiculous heights creating a high priesthood of statistical know-how that is largely unnecessary.
Variation isn’t that important anymore
In Ideavirus, Godin observed that 20 years ago, the top 100 companies in the Fortune 500 made something that you could hold in your hands. Today, only 32 make things you can hold. The other 68 traffic in ideas. That means that all of the emphasis on analysis of variation, which is so important in manufacturing, is largely unnecessary and overkill for two-thirds of U.S. businesses. Maybe it still matters in China or Malaysia, but not here. Forcing everyone to climb that pinnacle of knowledge to join the elite priesthood is a foolish waste of time and money.
The long tail of methods and tools
Like any other methodology, a handful of methods and tools will solve most problems. But there’s a long tail of methods and tools that will be used more and more rarely. Teaching people all of these methods and tools is a waste of time. From a lean perspective, it's overproduction.
Whether businesses are involved in manufacturing or services, all are transaction-based, meaning they rely on orders, invoices, purchases, payments, and so on. The only tools you need to know to handle these kinds of problems are control charts, Pareto charts and fishbone diagrams… in that order.
But aren’t control charts complex, you might ask? Not if you don’t make people learn all of the formulas for every kind of chart. You don’t need to know the formulas; you just need to know how to read the chart.
Aren’t control charts hard to create? Not with modern SPC software. PC-based software will handle the math and draw the graph. It will highlight all of the out-of-control signals so that you can get on with doing your job. Training people in control chart formulas is a waste of time.
But don’t you have to know what chart to choose, you might ask? I’d say yes, when you get more confident, but not initially. I found that after hours of teaching people how to navigate a control chart selection tree, they still weren’t confident picking a chart. That’s why I added a Control Chart Wizard to the QI Macros SPC Software. It will look at your data and pick the best chart.
Alan Kay, a well-known techno-wizard, said that a technologist’s job is to figure out how to make technology so cheap, easy to use, and ubiquitous that anybody can use it and it propagates around the world into every possible niche. Most of the user interfaces for today’s SPC software force users to clean their data, get it into columns, and tell the software everything it needs to know before it will draw a chart. Software like QI Macros takes whatever data is selected, cleans it up, reorients it and draws a chart.
Make it simple, then it will stick
Tom DeMarco, one of my favorite software authors, said that “making something complex seem simple is a huge intellectual feat.”
While it’s fascinating to watch intellectual battles over statistics between Forrest Breyfogle and Donald Wheeler, as we saw recently here in Quality Digest, it gives the average employee little hope. If these two can’t agree, they think, how am I going to be able to defend my analysis? Make no mistake about it; if someone confidently challenges the average employee’s choice of chart or data, they end up fighting an unwinnable battle. This is another one of the ways that we hamper the spread of lean Six Sigma, by making people feel stupid.
All we’ve done with Six Sigma is take something simple and make it annoyingly complex, which prevents it from propagating around the world and into every possible niche. This is what brothers Chip and Dan Heath, in their book, Made to Stick (Random House, first edition, 2007), call the "curse of knowledge."
Shame on us.
Isn’t it time to start focusing on the essential lean Six Sigma methods and tools so that we can take everyone on a picnic up a hill, so that someday soon, they will all be able to climb the pinnacles of quality?
Comments
It's All About Marketing!
Jay,
I think it goes beyond making Six Sigma complicated. It's been my experience that management waits until the 11th hour to address an issue that they've been aware of for quite some time. It's only when they're forced to address the issue that they consider Six Sigma, but only if it delivers a quick miracle otherwise they're not interested.
I've often marketed the 'low hanging fruit' concept to management as 'Rapid Six Sigma'....this way they experience positive results and are more willing to sign on for the duration of the project....it's all about marketing.
Sandra Gauvin
http://CurrentQuality.com.
a dirty secret of six sigma...
there is not one body of knowledge or even a universally agreed upon process for DMAIC (Six Sigma process/product improvement methodology).
For example, in my organization, Champions & Green Belts are given a lot of training on the LOGIC of DMAIC problem solving and very little statistics training; Black Belts get the LOGIC & the usual statistics training.
This gives me Champions who can review projects (and remember how to use DMAIC after training is over), a large population of potential Green Belts (who must lead a project) and Black Belts who are approached to be internal consultants (to keep spreading the culture while they work on their own assigned projects).
Why isn't everyone doing it?
Jay,
Great topic. I agree that most people still don't really understand Lean OR Six Sigma. That's one of the reasons Quality Digest published a paper that I wrote just a month ago about using Lean and Six Sigma to improve 'your airport experience'. It is a real problem that we have all experienced. I just showed how you can apply some simple tools to improve the time and reduce the variation in the trip.
The goal is to use the tools that are needed, when they are needed and keep it simple. Of course you still need to make sure management understands the philosophy of Lean Sigma; that's still critical in a rollout success.
Last, I have facilitated some of my best successes with teams when I DIDN'T actually give the team the name of the tool I was using. (my favorite is an FMEA). But once the team sees the result I let them know, 'oh, by the way you just performed a Failure Modes Effects Analysis'. It's great when their response is - 'really? that wasn't that hard'.
Six Sigma based on Nonsense
"Why isn’t everyone doing it?" My answer:
Because an increasing number of people have realised that six sigma is based on nonsense.
Read my articles:
http://www.qualitydigest.com/inside/six-sigma-article/six-sigma-lessons…
http://www.qualitydigest.com/inside/six-sigma-article/six-sigma-lessons…
Statistics battle between Forrest Breyfogle and Donald Wheeler
Thanks for an inspiring article, Jay!
I want to appeal Forrest Breyfogle and Donald Wheeler to get-together over beer and come-up with a limited number of bullet-points on statistical analysis that they both agree. This would be a great value-added activity for the statists i.e. people who apply statistics in their day-to-day jobs but are not trained statisticians.
Cheers!
Guru
(Gurbachan Chadha)
You sir are a heretic. Six
You sir are a heretic. Six sigma is the best thing since sliced white bread. It has done everything but cure cancer. Imagine if Ed Deming and Joe Juran were Black Belts, what a wonderful world this would be.
Seriously, this is the best Six Sigma article I have ever read. It exposes the fact that the emperor has no clothes. Most bb's and gb's I have met know enough to be harmful and not much more. The freshly minted bb with his suitcase full of statistics typically does more harm than good. Give me an experienced problem solver with detailed knowledge of the 7 tools, some Taguchi and most importantly the ability to understand what to change and how to change it over an army of black belts with their heads up their Minitabs anyday.
Great Article!
I totally agree Jay,
As we move more toward applying these principles in transactional environments in service industries we are faced with a couple of fundemental differences from manufacturing environments:
1. People are less likely to have fundamental statstics knowledge in service industries.
2. Transactional environments are much less likely to be as far down the road on their continuous improvement journey then manufacturing industries, and because of this, there is much less need to use complicated statistical tools to have a dramatic effect on results.
I have been working in the transactional world for 6 years now, and most transactional processes are so broken that basic statistical and lean tools are all that is needed to have make a big difference. It will probably be many years before the processes get up to 4 sigma levels - let along 6 sigma levels.
The challenge is how do we make process improvement simplified and easy so that those people who have never been exposed to quality tools realize the benefits and begin their journey.
I think your mountain hiking analogy is perfect. The answer is to take it slow and build their confidence.
marty yuzwa
Add new comment