I have nothing against six sigma per se. We used it at Danaher, Wiremold, and many other companies I was involved with in the private equity business. And while there are undoubtedly many ways to think about six sigma, I have always thought of it as a form of statistical process control (SPC), which is invaluable for finding and smoothing out variation in processes. I found SPC particularly helpful when dealing with very tight tolerances or processes with multiple inputs like temperature, flow, or time, where you couldn’t really see what was going on.
SPC was helpful to me when I was a group executive at the Danaher Corporation. We used it at Jake Brake (one of my group companies that made engine retarders for diesel trucks) when it had problems holding tolerance on parts coming off its screw machines. This lack of control created many rejects when we tried to assemble the final engine brake, which in turn was causing many defects. So, we decided to use SPC to try to eliminate the problem.
Soon after starting our SPC experiment, we started working with the Shingijutsu consultants from Japan. Specifically, my colleagues and I worked with three former Toyota executives who had spent their careers at Toyota, the last ten years working directly for Taiichi Ohno, the father of the Toyota Production System. They looked at what we were doing with SPC, kind of laughed, and just said “no” — as in “no good.” When we asked them why, they wouldn’t tell us, indicating that we would not understand, so why should they waste their breath.
… what alarmed me was seeing companies use six sigma as their primary cost-reduction approach — and, in fact, a core part of their strategy.
Instead, they showed us how to approach the tolerance problem from the lean point of view, which involved rethinking virtually every assumption we had made about how to do the work and improving the complete system rather than chasing down “point optimization,” or random improvements. We learned that controlling the setup and having fixed stops to maintain the tolerances and 100% inspection before a piece could move forward eliminated the problems. This approach also eliminated all the work we were doing to collect and analyze all the data, which was nothing but extra expense and therefore waste. We, in effect, eliminated the problem at the source.
Addressing this tolerance problem was a great learning experience. As I became more immersed in lean over my career, it became clear that occasionally using SPC could be helpful. However, what alarmed me was seeing companies use six sigma as their primary cost-reduction approach — and, in fact, a core part of their strategy.
Understanding the Limitations of Six Sigma
After I left General Electric to join Danaher, CEO Jack Welch got on the six sigma train in such a big way that it became almost a religion. And, of course, if Jack thought it was a smart thing to do, it wasn’t long before many other companies copied GE. I always thought this was a huge mistake for GE and a waste of their tremendously talented workforce.
The problem is that while you can realize some gains (admittedly, sometimes significant gains) from six sigma, this approach only goes after 10% to 20% of what is possible. That means you can have thousands of successful six sigma projects to brag about without much effect on the bottom line.
… while the six sigma approach fixed one of 20 steps, the lean approach addressed them all and delivered a lot more value to the customer.
Why? Let’s say you have a product line that has 20 value-added steps. Maybe one of the steps is a bit of a bottleneck, making it a perfect candidate for a six sigma team. You then focus exclusively on this step. Because you have smart people, you will improve this process and have a great six sigma story to share. Nice, but what about the other 19 steps? You still have those. They may be in different functional departments, but they still exist. So, on balance, you didn’t get all that much gain.
On the other hand, if you approach this with a lean mindset, you will look at all 20 steps from the beginning. First, you will take them out of their functional departments and line them up next to each other in a one-piece flow line. Then, you will look at the takt time, which will, in turn, help you see the waste in all 20 processes. As you improve each step, you will get enormous gains in productivity, quality, scrap reduction, lead time reduction, and customer service. So, while the six sigma approach fixed one of 20 steps, the lean approach addressed them all and delivered a lot more value to the customer.
Choosing the Easy Way or the Best Way
So why not just use the lean approach? You would think folks would want to run their businesses in this proven superior way. The answer is that lean requires a drastic change from the traditional way of running a business while six sigma does not. In the example I just cited, the lean team would go into the functional departments and rearrange all 20 machines into a one-piece flow line. This type of radical change is terrifying for the traditional manager. These machines haven’t moved since they came into the plant 15 to 20 years ago. Moving them will be disruptive and cost a little money.
And then, of course, there is the universal question, “But what if this doesn’t work?” Simply seeking six sigma gains involves none of these tough challenges. Everything can stay essentially the same, and the six sigma team can just work on the bottleneck process in the functional department where it lives. There is no disruption, no moving costs, and no worries about it not working. Six sigma, for most companies, is the hands-down choice. Whew, they avoided that lean nonsense.
So why not just use the lean approach? … The answer is that lean requires a drastic change from the traditional way of running a business while six sigma does not.
Six sigma also fits much better with the safe, studied way traditional companies solve problems. When faced with any serious issue to consider, most companies form a team that will meet once per week to study the situation and develop a plan. This team usually is a group of salaried employees who gather and analyze data. (You can easily see how a group of salaried six sigma experts could be the team members in this case.) Because not everyone can make every meeting, it might take two to three months to come up with and agree on a plan. Great. Now management has to approve this plan — another month.
The lean approach is quite different. It will also start by forming a team — but one comprised equally of hourly and salaried employees who will be taken off their full-time jobs for one week and given a very aggressive set of targets to meet by the end of the week. For example, they might start moving equipment by the afternoon of the first day. As a result, they will have a new line up and running by the end of the week with greater than 50% reductions in floor space, inventory, staffing, defects, and lead time.
Evaluating the Value of ‘Belt’ Certifications
And, what about black belts and green belts and yellow belts? Aren’t they important? My response is, “Look, you’re running a business not a karate class,” which always shocks people — maybe because it’s true. Companies spend a lot of time and money (up to several years) training a small select group of salaried employees to do SPC. Usually, they participate in six sigma study teams and then just return to their regular job. They may receive a nice certificate and a colored belt, but after all the training time, cost, and awards ceremony, they rarely end up doing six sigma full time. I prefer the lean approach of forming a Kaizen Promotion Office (KPO) staffed with your highest potential employees. Their full-time job is organizing, running, and following up on kaizen events. They learn lean by doing lean in a hands-on approach under the mentorship of an outside lean sensei. A good rule of thumb is to keep them in the KPO full time for about two years and then move them on.
‘Look, you’re running a business not a karate class.’
Six sigma enthusiasts have told me black and green belts are important ways to recognize the extra training given to their selected salaried employees. OK, I get the idea, but what about just promoting them? We pushed our best and brightest through the KPO so that they could become lean experts. Then, after two years, we promoted them to run something, like team leader, plant manager, or something else. That always seemed more important to our business than giving them a certificate and a belt they could keep in the drawer.
I’m sure I’ll get some negative feedback for this Ask Art. That’s OK. I’m just trying to point out that six sigma is comfortable because not much has to change with it. But unfortunately, that includes your results. On the other hand, adopting lean is very uncomfortable — until you see the results. You choose.