Used properly, Six Sigma can be a powerful process improvement methodology. By using data analysis to help determine how a process works, it is then possible to apply changes to that process where they will do the most good. The ultimate aim of a Six Sigma process improvement is to modify the process to the point where it will consistently deliver a product which is of an acceptable quality.
While Six Sigma relies on the use of various statistical tools, it is not necessary to have an in-depth knowledge of statistics to understand how it works1. However, it is important to understand that Six Sigma relies on data-based analysis to determine how to reduce variations in the process, and so ensure consistent delivery of quality output.
Why is Consistency Important?
Although a process which consistently fails to deliver is not good, a process whose delivery cannot be predicted, but which is not good enough, is even worse.
Generally speaking it is easier to improve a process which is stable than one which is erratic. For example, it would be easier to improve the handicap of a golfer who consistently hit their drive shots the required distance, but who always ended up twenty yards to the right of the fairway, than one whose shots sometimes went long, sometimes short, sometimes left, sometimes right.
'Quality' has a specific meaning in Six Sigma, being defined as 'conformance to requirements'. In Six Sigma there is no 'good quality' or 'bad quality', the output either meets the requirements, or it does not. Individual product output which does not meet the requirements is referred to as a 'defect'.
What Does 'Six Sigma' Mean?
Sigma is a measure of variation from the expected or optimal result. The higher the sigma value, the less the variation: a 'Six Sigma Process' is one which delivers to the process requirements with less than 3.4 failures per one million executions - ie one which produces the required result 99.99966% of the time2.
Who Invented 'Six Sigma'?
Motorola owns the copyright of the term 'Six Sigma'. The application of standard statistical analysis to process improvement was developed by them in the early 1980s, and 'released' as a formal methodology in the mid 1990s. It was further developed by General Electric and is now used by many large corporations.
Who Uses Six Sigma?
Six Sigma is used in large scale manufacturing industries, such as automotive, aircraft and semiconductor manufacture, where the continuous delivery of product 'within specifications' is essential to retain business and maintain profits.
It is also widely used in health care, where a 'product defect' could result in a loss of life; and in service industries where there is a requirement to consistently deliver a service within the terms of a formal contract to a large number of customers.
By far the most widespread application of Six Sigma in the Service Industry is in Call Centres, where the repetitive nature of the incoming calls, combined with the requirement to deliver consistent and timely responses to these calls, can be managed in the same way as a physical production line3.
How is Six Sigma Different From Other Quality Processes?
Six Sigma uses statistical tools to identify and remove sources of variations in a process, and optimise process performance. Although the process improvement principles used in Six Sigma are much the same as those used in other methodologies, Six Sigma focuses on measuring and managing the performance of components of the process during execution, to maintain delivery to the required standard, rather than measurement of the output to identify failures after the event.
The Phases Of A Six Sigma Project (DMAIC)
All Six Sigma process improvement projects are broken down into five phases, which are: Define, Measure, Analyse, Improve and Control.
In this phase, the level of process improvement required is defined and, in most cases, an estimate of the current process capability is provided.
Process capability and requirements must be defined in standard terms, the most common of these being 'DPMO' (Defects Per Million Opportunities), 'Sigma' (which is a measure of standard deviation from the mean), or 'Capability' (percentage which meets requirements).
In the Measure phase, the process is 'mapped', and measures taken to determine the current performance and scope of variation. This is most important since in many cases the perception as to the process capability is based on incorrect assumptions, and/or incorrect measurements. Also a process which has wide variations in output is, by definition, not 'in control', and so is very unlikely to be able to consistently deliver to any stated requirements.
Analysis is then carried out, using specific statistical tools, to determine process exceptions and the optimal operation of the process.
During the improve phase, steps are taken to modify the process inputs (the Xs) and operation such that the actual deliverable (Y) is optimal, or at least capable of delivering to the stated requirements.
During this phase, proposed changes are tested to confirm that the actual result of the change will be as expected, before they are implemented4.
Technically speaking, 'Control' is not a phase, in that it continues for as long as the process operates. However this is a vital part of the Six Sigma methodology, as it ensures the requirements of the process continue to be met. During Control, regular measurement of the process is carried out and adverse trends are identified and corrected. Ideally the process is brought back on-track before capability falls outside the requirements. This is made possible by an understanding of how the various components of the process (the inputs) contribute to the final deliverable (the output), and by identifying trends which would, if not addressed, cause an unacceptable level of defects.
An Extra Phase: Leverage
Many large organisations apply a sixth phase to their Six Sigma projects, called 'Leverage'. This phase determines where improvements applied to an individual process can be applied to like processes in other areas. For example if a multi-national company uses similar processes to supply call centre services from multiple locations, it is likely that the process changes which improve the performance of one location could be applied to all other locations with little or no customisation5.
What are 'Green Belts', 'Black Belts' and 'Master Black Belts'?
Green Belts have been trained in the basic principles of Six Sigma, and are generally those people who are involved in the day to day operation of the process. During the Define and Measure phases of the project they help gather data on how the process operates6.
In the Control phase, they are usually responsible for monitoring the operation of the process to ensure timely identification and correction of negative trends.
The Black Belt manages the process improvement, is responsible for performing the validation and analysis of the data gathered in Define and Measure, determining what process improvements need to be made and implementing them, testing the effectiveness of these improvements, and comparing the results with the expectations from the analysis. Finally they are responsible for delivering control mechanisms to the process owners.
Training in the Six Sigma methodology usually takes the form of four training sessions, each one week long and approximately one month apart. In order to become a 'Certified Six Sigma Black Belt', most certifying bodies require both course completion and successful review of one or two completed projects. These reviews are usually carried out by a Six Sigma process expert or 'Master Black Belt'.
Unfortunately there is no globally recognised certification body, and not all corporations are as rigourous in their reviews as they should be7.
Master Black Belts
A 'Master Black Belt' is usually someone who has extensive experience in the practical application of Six Sigma, and who acts as a coach or mentor to Black Belts. Usually it is a MBB who performs the analysis of the work done by a Black Belt to determine whether they are worthy of 'Certification'.
Why Six Sigma Doesn't Always Produce the Desired Result.
Many Six Sigma projects fail to deliver the breakthrough improvements expected. This can be for any number of reasons, but the most common is lack of rigourous application of the methodology. Six Sigma process improvement projects can take a considerable amount of time to execute, however cutting corners or making assumptions which are not supported by appropriate data analysis can often lead to disaster.
Other common reasons for failure of a Six Sigma project are: poor definition of requirements during Define, failure to identify all of the contributing factors which affect the quality of the output during Analyse, premature application of 'Improvements' before they have been properly tested, and/or failure to maintain 'Control' after the Improve phase has been completed.