Knowledge base
What is Six Sigma

What is Six Sigma


Piotr Ukowski

Achieving 99% effectiveness in action is a fantastic result. Is that so? 1% of errors means, for example, at least 4 aviation accidents per day at major airports or 200,000 wrongly prescribed drugs per year ...

The solution is to be a methodology, and in principle the philosophy of action – Six Sigma. The concept itself is derived from statistics. Sigma denotes the standard deviation of a variable and lets you measure quality, i.e. determine how good or bad the process is – how far the results deviate from the expected value and how many of them are within tolerance limits. One Sigma Level means 700,000 defects per million opportunities (DPMO), i.e. ... a catastrophe. 99% effectiveness of action ensures 3.8 Sigma Level. Six Sigma is just 3.4 defects per million (effectiveness 99.9997%). 1)

Statistical roots of Six Sigma are very important, but now, as I have mentioned, this methodology has become a management philosophy and is not only limited to managing production quality. The origin of the solution is simple – an enterprise’s operating costs must be reduced. It has been observed that a huge part of the expenses includes detecting and removing defects. These are direct costs (e.g. warranty repairs), but also losses due to the customers’ dissatisfaction, bad references, and a decrease in reputation. Six Sigma “observes” very revealingly that it is better to manage the problem than to eliminate it. It insists on detecting potential causes of errors and removing them before a malfunction occurs.

In other words, Six Sigma strives to ensure, with almost 100% certainty, that what is planned will be realized.

Another important assumption of this methodology is that, in order to improve something, knowledge on the current level and on the level desired to be achieved is required, as well as its presentation in the form of a measurable value. “If you cannot express in numbers what you want to say, you probably have no idea what you are talking about.”

From here, it is not far to the basis of implementing Six Sigma DMAIC 2):

  • Define

  • Measure

  • Analyze

  • Improve

  • Control

It is clear from the above that the implementation of Six Sigma is a continuous process lasting months, and often years, divided into dozens of tasks carried out in different teams and different places of the organization. It is the continuous improvement of the enterprise by developing its elements.

But let's go back to DMAIC.

  • Define. The basis is the proper definition of the problem. It is not about identifying the effects, but the process that is causing the defect. Precision is very important; it allows you to move on to the next stages.

At this stage, we need to choose the problems/ processes to deal with. This is another key decision. Effort must be focused on the improvement of elements which will ensure the greatest profit. How do we choose these elements? The ABC method supported by a well-built data warehouse can be helpful.

  • Measure. What do we measure? Many things. The number of mistakes but also the number of chances to make them. We need to know what happened and what was possible – we will recognize the scale of the problem. We need to compare ourselves with the competition – what are our advantages and what are they better in. Numbers give clarity, although their preparation is not easy (from experience it can be assumed that this phase lasts and costs two times more than it seems at its planning stage...).

  • Analyze. We have defined processes, we have numbers – it’s time to analyze them. Knowing what we have and what we could possibly have, we can calculate how much we will gain on improvement. If the profit is worthwhile, we go deeper – where are our mistakes coming from and how can we improve them?

  • Improve. It's not about the trivial “we have to try harder”. With defined and numerically described defective processes, we can identify specific areas for improvement. At this stage, it is necessary to involve very experienced, specially trained staff who know the business processes and the Six Sigma methodology well.

  • Control. There are two things involved. First, verification of the effectiveness of the undertaken actions, i.e. what we have gained, how much we have earned and what are the new values of our indicators. Second, continuous, active control of improvements to maintain the improved quality.

Six Sigma philosophy has been successfully implemented in large corporations (Motorola, General Electric). It assumes that actions should cover all areas of the enterprise. This statement is quite risky and it can only be used in very mature organizations. Additionally, a tendency for perfection often kills innovation which needs a much larger margin of error and a field for taking a risk. Therefore, implementation should be carried out very sensibly – special roles have been developed to ensure the effectiveness of actions.

First and foremost, control and support must be provided from the top – without clear actions of the management board, Six Sigma has no chance in the enterprise. Second, you need to invest – training of staff (the so-called Black and Green Strings) who will successfully implement the methodology is expensive and time-consuming (several months of training). You cannot expect a quick return on the investment; the first effects occur at the earliest after 6 months, more concrete ones after one year.

Six Sigma is triumphant. Multibillion profits are calculated from its use, and it is already present in Polish enterprises. Will it work for every company? Certainly not. The decision on its implementation must be made very carefully and be supported by appropriate training. Nevertheless, it can be treated as another “toolbox” – one can choose the elements that best fit into the company’s organization and give the highest return on the investment. Combined with controlling methods, Balanced Scorecard, it can give great results.

1) The inquisitive reader will notice that the figures indicated here deviate from the strict mathematical calculations. However, experience shows that most of the controlled processes do not have a perfectly normal distribution and over time there is a displacement of about 1.5 Sigma, hence the apparent divergence.

2) There is also one more expansion of this abbreviation: “Dumb Managers Always Ignore Customers”… It also accurately describes the idea of Six Sigma putting the Customer at the center of the enterprise’s interests