Sigma Level

What is Six Sigma level? It's a Lean Six Sigma metric that measures the error rate of a process, based on the DPMO estimate. The Sigma Level estimate is a long-term estimate of the process, including a potential 1.5 sigma shift that could occur over longer periods of time. This 1.5 Sigma shift is an "industry-standard" estimate of process Sigma Levels, as initially developed by Motorola and subsequently adopted throughout general industry applications.

A word of caution with respect to the use of DPMO and Six Sigma Level as lean six sigma metrics. There is a problem with a DPMO focus, and that is with the definition of an opportunity. While we can usually reach consensus with our customers on the definition of a defect, there is less clarity with the concept of an opportunity.

Consider a simple tax advice hotline. In this case, we define an opportunity as a single customer contact, regardless of the number of questions associated with the contact. If we instead define opportunities based on the total number of questions raised in each contact, the increased number of opportunities will reduce our DPMO for a stated number of errors. In this way, we will tend to think we are operating at a higher Sigma level.

This dilemma occurs in most processes. For example, a single satellite may have hundreds of main components and sub-assemblies, and thousands of circuits. If we define opportunities on the circuit level, we will certainly reduce our DPMO for a stated number of defects.

On the surface, higher Sigma levels sound like a good thing. Unfortunately, artificially inflated sigma levels do not lead to higher levels of customer satisfaction or profitability, which must be our ultimate objective. Unfortunately, there will always be some who will try to gameÂ the system in this manner, which of course detracts from all of the other Six Sigma programs that try to honestly estimate their customer satisfaction levels.

Since DPMO calculations can be misleading, we need to use them with discretion. Often we find that DPMO is still quite useful for internal measurements, particularly at the project level, once we realize its limitations and adequately define its calculation.

There are a few suggestions that can help:

We should classify opportunities in customer terms. How would a customer define an error, and how is this error delivered Â to the customer? If we look at the point of sale, or where the customer receives the product or service, then we are closer to understanding an opportunity in their terms. The customer does not care how many ways we can mess up (in other words, all the potential opportunities); they generally will look at the transaction as a single opportunity.

Therefore, an opportunity should be related to a transaction that results in customer appraisal.

Secondly, we should focus on only those characteristics or criteria that are truly critical to the customer. We can usually narrow this down to just a few criteria for a given transaction. It is not uncommon for cycle time to be one of these critical few. Countless non-critical criteria can be excluded.

Because the DPMO can be easily misstated, it is often a poor lean six sigma metric for management use. In fact, real criticism of Six Sigma program impact can be leveled against organizations that focus on DPMO at the detriment of other more meaningful metrics, such as financial benefit.