Breakthrough. It sounds good. It even sounds like something everyone should support. The problem, however, comes when people are asked to define breakthrough. There is almost always a slight pause, or in some cases a long one, that says they are not certain what it means.

To understand what breakthrough means and why it is an important part of any continuous improvement program, it helps to take a look at the history of the term in regards to Six Sigma and how the concept fits with another familiar Six Sigma idea: control.

Looking Back at Juran

In terms of the quality discipline, Joseph Juran was the first to publish a cognitive effort around the concept of breakthrough in his book Managerial Breakthrough: The Classic Book On Improving Management Performance (McGraw-Hill) published in 1964. He defined breakthrough as “dynamic change.” By means of contrast, he defined control as “lack of change; maintaining status quo.” In the context of Juran’s definition, breakthrough and control are mutually exclusive; it would be impossible to have one at the same time you were involved with the other.

The concept of control is best understood through control charts. Control charts are a tool that allow practitioners to collapse process variation around some measure of central tendency and separate natural variation from an assignable cause.

If reducing variation is control, how do you achieve breakthrough? Why would you want breakthrough since it seems to mean the opposite of control? Figure 1 is similar to the illustration that Juran published in Managerial Breakthrough and shows the benefit breakthrough can bring.

Figure 1: Representation of Control and Breakthrough
Figure 1: Representation of Control and Breakthrough

The illustration does not show control and breakthrough as opposite things as much as it shows them as elements of an improvement system. There is a process that is in some state of control; it is driven to a different target and control is reestablished around the new target. The new state is within tighter control limits, meaning the effects of different variables were studied and practitioners determined which variables affected the mean and which affected the variation. This sorting process allowed the process to be reestablished to an improved state. Although the two definitions (breakthrough and control) initially appear diametrically opposed, when they are used together they become a basic roadmap for achieving breakthrough results.

The Six Sigma methodology aligns with what Juran published decades earlier. The shift in Juran’s illustration is a mean shift. In the Six Sigma methodology this occurs in the Improve phase. Juran’s illustration shows control being reestablished once the mean shift is accomplished. The Six Sigma methodology has Control as a final step to institutionalize the breakthrough that took place.

Why Make a Breakthrough

Why would taking a process through breakthrough be a desirable thing? Taguchi’s loss function can provide that answer. The idea is that there is an optimal place, or target, to operate every process. The loss function states that as a process departs from that target there is a “loss imported to society.” When using the concept that Juran illustrated or the methodology of Six Sigma, the process is consistent with Taguchi’s idea – drive the process to an optimal target and reduce variation around the target. The critical element is establishing the target that satisfies both the needs of the customer and is the most financially viable target for a producer.

Selecting the Right Target: A Breakthrough Example

The following case study illustrates the breakthrough concept. There is a plating process that sputters gold. The customer specification is between 900 angstroms and 1,200 angstroms (Figure 2).

Figure 2: Customer Specifications for Plating Process
Figure 2: Customer Specifications for Plating Process

The product as defined by the customer is acceptable within these limits; therefore, the most financially rewarding target for the manufacturer is a process with a mean equal to 900 angstroms and a standard deviation of 0.

The current process is targeted low so the mean is toward the lower specification limit. Because the variation is large, the customer is receiving out-of-specification material and they are unhappy. The manufacturer is also dissatisfied with the reject rate and the customer’s complaints (Figure 3).

Figure 3: Current Production Data
Figure 3: Current Production Data

The first step must be to protect the customer, so the target of 900 angstroms is not practical. It would worsen the dissatisfaction of both the customer and the manufacturer.

After applying Six Sigma theory, in its pure state, the new target is six standard deviations inside the lower specification limit; this is a breakthrough change (Figure 4). The customer is protected and satisfied; however, the new target is not the optimal target of 900 angstroms. Based on Taguchi’s loss function, there is a loss in terms of additional gold plated on each piece and that increases material cost for the manufacturer.

Figure 4: Shift in Mean Six Standard Deviations
Figure 4: Shift in Mean Six Standard Deviations

The next step is to reduce variation around the mean. It is that decrease in the standard deviation that allows the mean to be shifted toward the lower specification limit or the true target. A shift of the mean, another breakthrough step, toward the lower specification limit with the smaller standard deviation allows the same level of protection for the customer and allows the manufacturer to reduce its cost (Figure 5). There is no disadvantage to either the producer or customer in this process.

Figure 5: Second Breakthrough Improvement
Figure 5: Second Breakthrough Improvement

Incorporating Breakthrough

Establishing a continuous improvement program that focuses on breakthrough introduces the mean shift as a key approach to success. When practitioners focus exclusively on variation reduction, or control, they may be optimizing a process on a suboptimized target and end up leaving huge benefits lying on the table.

Regardless of the tool set used, when building a continuous improvement program, it is important to incorporate both control and breakthrough methodologies.

About the Author