It is the first week on the job, and you find yourself looking over the various processes of the new company. Production line 1 has a sigma level of 3.4, while production line two has a sigma level of 4.8. What do these numbers represent? Which line is doing better? 

You ask a supervisor while touring the production floor, but all you get in response is, “I really don’t have any idea what these numbers mean. I’ve been told higher is better for quality, but I don’t know why. It’s all Greek to me.” 

The last response elicits a half smile from you, but you walk away not really understanding this metric. What does sigma level mean?

An overview: What is a Sigma level?

Sigma level is a statistical calculation that takes short-term information regarding the defects per million opportunities (DPMO) of a process, factors in the inclination of a process to shift over time, and gives a level value score representing the modified DPMO with shift in an attempt to help determine if the quality capability of the process can meet customer requirements over time.

For example, 1-sigma represents a poor level of quality, with 691,462 defects per million opportunities over time, which demonstrates a failure rate of over 69.1%. A level of 6-sigma represents a high level of quality, with 3.4 defects per million opportunities. It is this sigma level that leads to the term Six Sigma, which is a philosophy of delivering near perfect products or services by eliminating variabilities that lead to defects.

Motorola engineers discovered that many of their operations were prone to shift 1.5 sigma over time. Assuming normal distribution and normal variation of the mean, the specification limits would need to be set at a particular sigma level to achieve the desired defects per million opportunities level. 

For example, if the long-term desired maximum defects per million was 6,210 defects per million opportunities, factoring in the 1.5 sigma shift over time, a specification limit of +/- 4-sigma would be required. The goal of many programs — a goal few would meet — was to reach a specification limit of +/- 6-sigma, or 3.4 defects per million.

The actual calculation for sigma level is commonly calculated by using a spreadsheet or online conversion webpage. Start by calculating the defects per million opportunities (DPMO) as it is currently understood. Next, find a sigma level conversion table to determine the sigma level. The calculation will factor in the 1.5 sigma shift over time and give you the sigma level for the process. This page will also give you defect percentage and yield values to expect over time.

Certain advanced understandings for sigma level can be found here. This article goes into additional details regarding the formula as well as more information on how the formula is designed.

3 benefits of attending to sigma level

Sigma level is often used as a way to understand the current quality capabilities of a process, how this level matches up with customer expectations, where possible areas of improvement may reside, and as verification that a process improvement project has had the intended effects.

Understanding your current process quality capabilities versus customer expectations

Remember that value is determined by the customer. Some processes can afford to have a lower sigma level, as the level of quality required by the customer is not the highest. The customer may not always be willing to pay for a process with the highest quality standards.

Other processes, such as those that can affect lives, must be capable of reaching a much higher sigma level. Measuring sigma level helps you understand the current quality capabilities of a process, but it also requires a company to ask if this level is set at the right level for what the customer is willing to pay, and for the safety requirements of the process.

Demonstrate lost capacity over time

Low sigma levels can be a great place to investigate lost capacity over time. Each time a defect occurs, it becomes a non-value added drain to the organization. Significant costs to material and labor used during defect correction can occur. When sigma levels baseline values are extremely low, it can be a major cost savings opportunity to address.

Verify improvements achieved by process improvement efforts

If a process improvement event addresses a key quality issue of a process, there should be an incremental improvement in the sigma level of the process. Measure the sigma level prior to and after improvement are in place to verify that the improvements are improving quality as intended.

3 sigma level best practices

  • Even if the quality level the customer wants is met, remember that any improvement past that point can still be a cost savings for the company. In contrast, there becomes a point where you get diminishing results from improving the sigma level, so be sure that these improvements don’t cost more than the cost of the defects they are addressing.
  • As with any metric used to determine future process improvement projects, be sure to address the areas that will result in the highest improvement to quality.
  • For the vast majority of companies, a sigma level of six should be the goal, but don’t expect to achieve it. It’s more important to understand how much your customer is willing to pay for quality — and that the desired sigma level can be met at that cost.

Frequently Asked Questions (FAQ) about sigma level

Does my process need to be at 6-sigma in order to be successful?

No. A level of 6-Sigma requires an incredibly low failure rate, which is a level of quality most processes don’t need to achieve to be successful and meet customer satisfaction. Remember that your customer determines the quality expectations, and most customers are not looking for near perfection — nor are they willing to pay the cost requirements needed to reach such a high level of quality. 

Many companies can actually compete at a 4-sigma level, which allows for 6,210 defects per million. Certain industries, such as health care and airlines, require higher sigma levels due to the fact that lives may be ruined or even ended by defects in the process. A 4-sigma level translates to an aircraft crash rate of 0.6%. With multiple millions of flights each year, a crash rate of over 5 planes a day is not something the customer would tolerate. These industries are always striving to get as close to a six sigma level as possible.

Is it a guarantee that my process will shift by 1.5 sigma over time?

No. While quite a few processes follow this trend, some processes will deviate by much less than 1.5 sigma over time. It’s also possible for a process to deviate by more than 1.5 sigma. Knowing the characteristics of the process are key to knowing whether you want to use sigma level as a business metric. Does the process follow the characteristics of a normal bell curve? Some non-normal processes are not good candidates for sigma level. 

How does Cpk factor into sigma level?

Cpk is a value representing how many standard deviations away the specification limits are from the process mean. It’s yet another tool for understanding whether customer requirements are met. It has a linear relationship with sigma level, with each sigma level being equal to one third Cpk. For example, a sigma level of three sigma would have a Cpk value of one.

Sigma level represents the quality and capability goals of a company for long-term customer satisfaction

Sigma level is an easy-to-use approach that factors in statistical shift of specification limits to match capability expectations with customer expectations. This value is understood across all industries, and it’s important to monitor it to be sure your company maintains quality over time.

About the Author