Pei-Chi Chiu　|　February 23, 2023

Six Sigma at GE (Part 2)　Six Sigma in Science and in Management (Part 2)
[Six Sigma at GE]

Motorola began promoting its 6σ improvement program in 1987, and even after it won the Malcolm Baldrige National Quality Award two years later, it still did not attract a great deal of attention from the business world.

Yet the 6σ program that GE re-launched at full force in 1995 drew enormous attention worldwide.

Why did GE’s 6σ program become so famous? Was it because of the all-out drive of GE’s legendary former CEO Jack Welch, or because of the astonishing results GE reported (taking 1999 as an example, the program delivered roughly US$2 billion in benefits), or because GE’s 6σ program was overhyped? Each of these is one of the important reasons.

Many people believe that 70%–80% of the credit for the success of GE-6σ should go to Jack Welch.

This view has some merit, but a number of other supporting practices were also required to make the program so successful.

1. The determination and commitment of senior management

Jack Welch’s determination and persistence were indeed a very important factor in the success of GE-6σ. This is also a prerequisite for any company to promote a 6σ program. Promoting a 6σ program requires more resources and training than introducing other management systems, calls for greater cross-departmental teamwork, and also demands outstanding improvements in (or transformation of) business performance.

It is therefore relatively difficult to drive forward, so beyond the strong determination and persistence of senior management, communication with all employees must be intensified to form a consensus, in order to win the cooperation, support, and action of the entire workforce.
In addition, major undertakings must be carried out, such as setting a vision, developing a direction, driving change, and effecting a cultural transformation.
2. Strong leadership from managers and ownership of the 6σ projects

In the way the 6σ program is designed to operate, the managers of each department are the providers or designators of the 6σ projects, and they provide the necessary resource support for the projects they supply.

Moreover, the Champions, Master Black Belts (MBB), and Black Belts (BB) all come from senior management.

They all need to play the role of good leaders, and because the improvement (transformation) results of 6σ projects must be very outstanding, departmental managers and senior executives all need to possess strong leadership capabilities.
3. A structured role design

The general business world and the management community tend to focus, when it comes to GE-6σ, on its operating steps DMAIC (Define-Measure-Analyze-Design-Verify).
While these operating steps certainly have functions that cannot be overlooked, the most important success factors lie on the organizational and execution sides.
To achieve good implementation results in these two respects, GE specially established the Six Sigma operating roles and infrastructure.
* Champions
They are the designators of the 6σ projects and also the sponsors, bearing full responsibility for the successful operation of the projects. They therefore need to provide financial support and to help the project teams solve major problems.

* Master Black Belts (MBB)
Master Black Belts can be regarded as full-time training instructors for the operation of 6σ projects, and therefore must possess sound quantitative techniques along with teaching and leadership abilities.

* Black Belts (BB)
Black Belts are the leading figures of the 6σ projects and the full-time executors of the 6σ projects. They lead the project teams and are responsible for executing the 6σ projects throughout.

* Green Belts (GB)
They are members who participate in the project teams led by the Black Belts. They are not full-time participants and therefore still have to carry the responsibilities of their original jobs.

GE is very willing to invest in promoting 6σ projects, and it places great emphasis on the returns on that investment, which is why it has achieved such good results.
From the CEO and Champions all the way down to the Green Belts, GE has a complete education and training program, and it is implemented thoroughly.
The CEO and Champions receive one to three weeks of education and training, focusing on how to formulate a vision and strategy, how to manage change, as well as leadership and communication skills.
The heaviest education and training load falls on the Black Belts, who receive at least three to four weeks of training, conducted in tandem with the operating steps DMAIC (or DMADV).
Usually each phase begins with three days to one week of training, followed by three weeks of execution. The training content mainly covers the way DMAIC operates, the statistical tools that need to be used, as well as project management, communication, and leadership.
As for the non-full-time Green Belts, they also receive about two weeks of education and training, the content of which likewise covers the way DMAIC operates and statistical methods.
GE-6σ places great emphasis on company profitability and development.
The promotion of TQM particularly emphasizes the pursuit of customer satisfaction, but in pursuing customer satisfaction it does not necessarily make the company profitable.
However, because Jack Welch was a professional manager, he had to make the company profitable, and so he had to balance customer satisfaction with company profitability.

Although any improvement item under GE-6σ is certainly driven by the voice of the customer and considers the critical-to-quality (CTQ) elements as the focal point,
the improvement (transformation) results are nonetheless measured by financial performance, because the improvement goals incorporated the original Motorola 6σ concept.
As a result, the magnitude of improvement is roughly tenfold or more, and at the very least several times over, and consequently the financial performance targets set for each 6σ item are extremely high.

Because senior management is directly responsible for the 6σ projects and provides the necessary resource support, the likelihood of achieving these performance targets is very high.
To make the personnel involved in the 6σ projects give their all, GE specially designed it so that 40% of the company’s year-end bonus is determined by the execution results of the 6σ projects.
Not only that, but how good or bad the execution results are also affects the promotion of the personnel involved.
These practices of GE’s are indeed very effective, and they have indeed made the business results of the 6σ projects very good. Since the program was launched in 1995, it has already generated billions of dollars in benefits.
In 1999 alone, the profit reached as high as US$2 billion.

Beyond valuing the company’s short-term profitability, GE-6σ does not neglect the company’s long-term development either, and so GE’s 6σ projects must be integrated with the company’s development strategy.
The company’s strategic planning connects to the 6σ projects; many of the company’s strategies rely on the 6σ projects to operate.
Therefore, we cannot simply regard GE’s 6σ items as merely improvement or process-reengineering tools, like QCC or re-engineering.
GE-6σ not only pursues customer satisfaction and pursues company profitability, but is also a part of the company’s strategic development. [Six Sigma in Science and in Management]

Generally, quality-control problems basically assume that the quality characteristic being controlled has a normal distribution. The standard deviation (σ), which expresses the degree of concentration of the distribution, and the mean (μ) are the two important parameters of a normal distribution; this is the basic understanding in quality operations. Therefore, a certain quality characteristic of a certain product can be measured as X, and under the assumption of a normal distribution X~N(μ, ). For the quality required by 6σ, the nonconformance rate should be 0.00198 ppm. Statisticians all understand that, scientifically, to check whether the assumption X~N(μ, ) holds, one must obtain hundreds of millions of observed values of X before this condition can be verified as usable—not to mention small and medium-sized enterprises, even large multinational corporations probably do not have such capability. As for the 1.5σ cited in the quality classification discussed by Motorola and GE, this performance can also be well understood by looking at it through the t-distribution. The ppm value reflected by a kσ system is not all that absurd. Therefore, the term “standard deviation” in Six Sigma should not be the objective; rather, if one wants to maintain a 3.4 ppm system, instead of striving for 6σ, it would be better to put effort into SPC. Because if SPC can guarantee detection of a 0.5σ shift, then a five-standard-deviation system can also produce a 4.3 ppm defect rate, which has the same effect as a 6σ system that needs to tolerate 1.5σ. What truly needs to be done is to work hard to reduce the standard deviation, improving the process with quality techniques such as SPC and DOE.

Every statistician understands that Six Sigma is not easy to verify scientifically as to whether it is usable, but no one will boycott this trend because of that. In a profit-driven market, any means that can create profit is a darling of the times, and rigorous scientific verification is no longer all that important.
Perhaps the fashionable term 6σ can methodically narrow the gap between the most senior level and the front-line execution staff with respect to quality goals, vision, concepts, definitions, and procedures, and one can see some data-based process improvement techniques. However, water can carry a boat and can also capsize it; with a correct understanding of the meaning behind statistical methods, and the correct use and promotion of statistical methods, perhaps problems can be solved easily without going to great lengths. Doing things right the first time may be even more powerful.