Posted by: knightbird | January 11, 2017

Are Statistics Necessary for making Lean Improvements


As a Lean Strategic thinker, I am not a fan of Six Sigma. While researching use of control charts for analysis of increases or declines in annual suicide rates over time, I ran into a lot of Six Sigma discussion. I have three observations I usually make, and I am sure there are more.

I am not a great statistician by any means. I struggle with any concept other than the basics. Becoming a belt requires some statistical training, and that limits the number of employees who can become process and quality leaders. When someone has the credential, the are no longer a part of the team, but become a supervisor. The Belt applies DMAIC in their projects and because of its complexity, are deferred to in the improvement process. Building a team, and teaching them the principles of Lean, are key to any Lean cultural implementation. By anointing Belts, the learning process is not shared equally, and the culture does not spread effectively. Lean is simple in my mind, see a problem, fix the problem. Use any means necessary, and to it now.

Most Six Sigma projects I learn about take time. My colleague at a federal agency told me that their Belt quota annually was 2 projects completed. Kaizen is slow, evolutionary change. Kaikaku is radical change, and generally attempted by Sensei’s guiding employees. When Sensei Nakai began working with Porsche in the early 1990’s, he did Kaikaku. He basically ordered change, and the team moved quickly. You can do that when you just lost $150 million. To practice Kaikaku eliminates the need for Six Sigma. Changing a process in manufacturing is easy and a Sensei can see the waste without the need for statistical analysis. When you have achieved a high level of Lean, then you can incorporate statistical measures into your processes for an early warning system. If you have a machine that must meet fine tolerances, it pays off to measure a small output of the product and construct a control chart. When the chart shows a series of downward measures of tolerance, it is a sign that the machine is wearing, and will need maintenance to return it to tighter tolerance. The measures and control chart can be set up by a statistician and employees taught how to maintain the measurement system and input the data for computation.

When you start implementing Lean, the correct way, you want to do it through Kaizen. It’s a teaching tool. When you run 3 concurrent Kaizen, as one of my organizations did, you maximize the value of a Sensei by bringing the teams together to develop a problem statement, create an A-3, map a value stream, walk the Gemba, gather data, and perform the many other tasks that contribute to a successful Kaizen. The teams separate to work on their own problem statement. I was never concerned, after I learned the power of Lean, about the results coming out of a Kaizen conducted by a Sensei. They happen and are far more successful than you ever imagined. And that creates one problem. Command and control leaders who have resisted are shamed by the extent of improvement achieved. The goal is not to improve the process It is to engage and educate your employees about continuous improvement. A well-managed first pass at a Lean event with any problem statement involving a process is always successful.

Very few Lean conversions are truly successful. Because my son lives in Independence, OR, I read the recruitment notices for Lean Managers and found one manufacturing company seeking a Lean Manufacturing manager. I applied for it in the hopes that I might be interviewed, without any expectation of meeting the qualification hurdles of 10 years in a manufacturing environment. The company has a great video of their production process on their website and I looked at it as a Sensei. I could see some areas where actual flow would be improved. However, most manufacturing processes are already efficient. Decades of copycat thinking (derivative) means that changes are forced on employees and they remain static until the next copycat change is found. The real benefits remain waiting to be discovered in all the other related processes, from Just in Time implementation to hiring, training, accounting, procurement, records management, safety and a host of other affiliated processes.

A good Sensei can educate the CEO and show them the benefits of a Lean Implementation done correctly. They can help mid-level managers understand how to achieve outstanding results with fewer resources. And they can energize and excite employees by training, coaching and mentoring. Outside of Kaikaku, a Sensei truly want the team to discover problems and solve them. And if they are not all solved at once, so what. We can’t have continuous improvement if we can achieve every improvement at one.

Implementing Lean isn’t expensive, and the payoff is rapid if you know what you are looking for. If you can figure out how to do the same amount of work with 50% fewer employees (a realistic target in my experience) you have huge savings, in salary, benefits, space along with tremendous productivity increases. And a good Sensei can conduct one event a week, and mentor hundreds of improvement events in a year. With Belts, it takes a huge training investment and certification by an organization that you meet the requirements for the level of Belt you are trying to earn. After the substantial investment, a Belt is marketable, and you generally must pay a Belt more after the certification, or they can move to a better paying job elsewhere.

Higher level statistics aren’t necessary for starting with Lean. There is no substitute for starting at whatever your skill level is. Dive in. Bring in a Coach or Teacher. If you are in distress and not sure whether you will survive, then by all means conduct Kaikaku. It’s like surgery conducted by an expert. But otherwise, practice prevention for bad stuff that happens to organizations by starting Lean.




  1. I agree with your sentiments. I wrote a post a few years ago “Lean: Out of Reach” that was premised on the seeming need for pre-requisite requirements (the belt) to participate in lean initiatives. Unfortunately, some even suggest that this is an inherent evolution in the lean process.

    I can appreciate the application of statistics and support reducing and eliminating variation in the process. I consider variation to be the greatest source of loss and waste. This extends beyond the traditional quality control methods of managing product quality. Statistics can also be applied to production data (variance between shifts), process parameters or other elements that affect the throughput of a given process or system.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: