Management
Management

Six Sigma Redefined

Applying Six Sigma can be deceptively intimidating.

November 7, 2013
KEYWORDS Praveen Gupta
/ Print / Reprints /
ShareMore
/ Text Size+
Trans

Six Sigma is one of the most popular and successful quality management tools around. The process improvement system, which was birthed at Motorola, has become famous for saving organizations money and helping them improve their quality processes. Yet, as the Six Sigma certification process has blossomed, it has also incorporated more room for error. Praveen Gupta, director of corporate quality at Prysm, a maker of large display solutions, and a longtime quality teacher, says that this is often because leaders and employees get lost in the statistics. Instead of focusing on statistical tools, we should concentrate on process knowledge and statistical thinking, he explains.

In a recent interview with Quality, Gupta went into detail about the right way to approach Six Sigma:

Quality Magazine: Why is Six Sigma so popular?

Praveen Gupta: The actual reason for the popularity is that there’s no other methodology that has received credit for saving so much money. Companies have reported savings in the billions of dollars. No other methodology comes close to Six Sigma for savings. Organizations will implement any methodology, if it improves quality or saves money. Six Sigma offers both. That is why many corporations like to deploy Six Sigma.

QM: Has Six Sigma changed?

PG: Initially, the intent of Six Sigma was to develop a methodology to achieve dramatic improvement quickly. To achieve dramatic improvement, we would need collaboration, aggressive goals and creative approaches.

The primary purpose of developing the Six Sigma methodology at Motorola was to become the best in everything it did. Prior to applying Six Sigma, people at Motorola were applying various quality improvement methodologies, but they were deployed functionally, locally and sub-optimally. Motorola customers were not experiencing improvements. They looked into it and developed the Six Sigma methodology, which is a collection of many conventional quality improvement tools all put together. In addition to the quality improvement tools, concepts of project management, such as project score, charter, etc., were incorporated to ensure desired results. So Six Sigma is more like a tool box with conventional quality improvement tools and a project management mindset.

Over the years, Six Sigma has changed. Six Sigma was originally started to be the best in everything one did. Initially, it was a six-step methodology, which later transformed into Define, Measure, Analyze, Improve and Control (DMAIC).  To institutionalize Six Sigma through in-house competency, training and certification programs such as Black Belt, Yellow Belt, and Green Belt were developed.

In recent years, as people received extensive training and implemented statistical tools, many Six Sigma practitioners were bogged down in the methodology and the statistical tools. As a result, many Six Sigma projects were not closed, and opportunities for dramatic improvement were missed. In reality, most people do not like statistics. But, many instructors thrived on jamming on statistics.

For example, a team is formed to solve a problem using the Six Sigma methodology, DMAIC. People go through the Define phase, where the challenge is to get the past data and establish a benchmark. It makes it difficult to prioritize projects and teams end up working on projects with an insignificant scope for savings. Moving into the Measure phase, many times it is difficult to get hands-on data for establishing a baseline and analyzing for variations and patterns. A Six Sigma practitioner finds a way to interchange data with knowledge, and knowledge with data.  People get lost in the Measure phase, in finding sources of data, or substitute with collective intelligence. Similarly, if one is not careful, one can spend a lot of time and resources in each phase in order to follow the methodology, but not solve the problem.

QM: What else has changed over the years?

PG: Initially, Six Sigma was developed as a methodology to solve problems. This included the use of the quality and statistical tools needed to tackle chronic or severe problems. Originally, Six Sigma was defined as a methodology to achieve best-in-class performance. Given its success, at one point Six Sigma was even stated as a way of doing business differently, or the DNA of a company. Over the years, the definition has lost its objectives and become more like a fact-based, decision-making process for process improvement. Now, practitioners are busy working on Six Sigma projects, but are unable to close them due to lack of improvement. Initially, the intent was to realize breakthrough improvement, but now it is being used to improve, even incrementally. Considering investment in Six Sigma training, many companies are unable to realize return on their investment because of difficulty in using statistical tools, incorrectly using them, or overly depending on statistics and statistical software, looking for correlation. Understanding of the difference between correlation and causation has blurred.

Another example is, when people are improving a process, typically they’re not reducing variability, instead they’re adjusting the process mean. If I were reducing variability, I must be redesigning the process, or doing something very differently, to see dramatic improvement. That’s why Six Sigma was considered as re-engineering of every process. If we are not seeing breakthrough improvement, we must not be re-engineering processes, or using the Six Sigma methodology as intended.

Instead of being a means to achieve performance improvement, Six Sigma has become the end in and of itself. Businesses are not there to be working on Six Sigma. Six Sigma is meant to work for business in reducing waste and improving profitability. In most companies that have implemented Six Sigma, they do not measure their level of Sigma. If not measured, it is less likely to be monitored, implemented effectively and sustained over time to create a culture of excellence.

QM: So what do you feel is the reason Six Sigma is becoming less popular?

PG: Reasons I have heard for Six Sigma losing its popularity is that it is not proving its worth. Projects are taking too long. Six Sigma is being applied on trivial projects, without potential for savings. Six Sigma training programs have lost their rigor. Also, the Six Sigma infrastructure of belts has caused loss of teamwork among belts, and belts do not correlate with the hands-on expertise in problem solving.

QM: What is your advice for those companies who could do a better job of applying it?

PG: Six Sigma is a comprehensive approach using a large collection of improvement tools. Instead of focusing on statistical tools, we should focus on process knowledge and statistical thinking. If we use DMAIC in conjunction with process knowledge, we will see better results. Statistical thinking is more important than statistical tools. Statistical thinking implies understanding the nature of the variation, or the difference between random- and assignable-sources of variation. Our ability to understand the nature of variation will determine whether we need to adjust the process or redesign it.

I have found 80% of Six Sigma methodology does not include statistical tools, and 80% of tools used in Six Sigma are non-statistical. However, the perception of Six Sigma methodology is that it’s all about statistics, and most people feel that statistics are difficult to understand. I find that this false perception makes Six Sigma less popular and discourages people in adopting the Six Sigma methodology.

QM: So you would encourage most prospective Six Sigma users to put less emphasis on the numbers and the statistical measurement and more focus on their own individual processes?

PG: In many companies, there is a lot of data, but it is ignored. In others, there is inadequate data collection and decisions are reactive in nature. If the data is available, we should use it effectively and maximally. While solving a problem, I emphasize the value of improving process knowledge more than getting busy with using statistical tools, but these tools are important and must be learned to answer unknowns about the process. In other words, problem solving is about gaining missing process knowledge. So, the No. 1 priority is to understand the process better.

Secondly, learn about statistical thinking before learning statistical tools. In the absence of a proper understanding of statistical thinking, statistical tools can be easily misapplied and valuable, limited corporate resources can be wasted.

QM: And what are some examples of those statistical tools?

PG: Design of Experiments (DOE), a popular tool among Six Sigma professionals, comes to my mind. I have seen people start solving problems with a DOE. It used to be that people would go through a Six Sigma Black Belt training over four months. During the last week they were given a heavy dose of statistics, such as DOE. In a long training, the tools taught in the final week will make a stronger impression than the ones learned in the first week. And DOE is a very powerful tool; however it requires process knowledge to conduct a DOE correctly. We forget that we need to use other, simpler tools to beef-up our process knowledge before we conduct a DOE, to confirm our understanding of process variables or prioritize potential variables.

We must understand that the Define phase has practically no statistical tools. We also need to understand that a problem well-defined is half-solved. I have also observed that many designed experiments are closed as inconclusive, which represents a lack of process knowledge, and [incorrect] use of DOE. Experts have said that 80% of the time must be spent in designing an experiment, and 20% in doing it. Reality is otherwise. Selecting a correct set of variables for an experiment is a critical step and requires process knowledge. When an experiment is inconclusive, or does not show a statistically significant change, it represents primarily selection of wrong variables due to a lack of process knowledge.

We must understand that statistical correlation does not mean causation. But the causation shows correlation. If we have process knowledge, we have a better chance of solving the problem, versus learning statistics and then solving the problem.

 Praveen Gupta, an excellence and innovation thought leader, currently works as a Director of Corporate Quality at Prysm, a maker of large display solutions using its patented laser phosphor display (LPD) technology. He has authored many books on Six Sigma and innovation, is the founding Editor-in-Chief of International Journal of Innovation Science, and co-editor of Global Innovation Science Handbook.  

Did you enjoy this article? Click here to subscribe to Quality Magazine. 

Recent Articles by Genevieve Diesing

You must login or register in order to post a comment.

Multimedia

Videos

Podcasts

 In honor of World Quality Month, we spoke to James Rooney, ASQ Past Chairman of the Board of Directors 2013, for his take on quality around the world.
For more information, read the ASQ Speaking of Quality column.
More Podcasts

Quality Magazine

CoverImage

2014 July

Check out the July 2014 edition of Quality Magazine for features!

Table Of Contents Subscribe

The Biggest Obstacle/Concern?

In the current economic and business climate, what is the biggest obstacle/concern to your job?
View Results Poll Archive

Clear Seas Research

qcast_ClearSeas_logo.gifWith access to over one million professionals and more than 60 industry-specific publications,Clear Seas Research offers relevant insights from those who know your industry best. Let us customize a market research solution that exceeds your marketing goals.

eNewsletters

STAY CONNECTED

facebook_40.png twitter_40px.png  youtube_40px.pnglinkedin_40px.png