Why Six Sigma: Introduction and DMAIC

Six Sigma is a concept you will have heard managers talk about all the time. You may think, well, this is a complicated procedure suitable only for big manufacturing corporations who can afford that entire infrastructure. How can it be applied to a smaller-scale IT Company like yours? After all, Six Sigma does need a special set of experts (called Black Belts), and you may be in no position to afford them.

Yes, Six Sigma is a complex process. Yet, time and again it has been shown to have helped companies reduce expenditure, improve efficiency and lead to greater customer satisfaction, which are the three goals of any for-profit organization, whether in manufacturing or services. It is doubtful if a company can continue not to implement Six Sigma and yet stay competitive, at a time when your competitors are actively adopting it.

A quick introduction to Six Sigma and its goals. A strategy pioneered by Motorola in 1986, it aims to achieve consistent and predictable results within any process, such that there are less than 3.4 defects per million opportunities (DPMO). That is, for every million repetitions of a process (say making a microchip), only 3.4 times does the product fail to meet the specifications.

You may ask, what is so great about Six Sigma. This is because every process has an intrinsic efficiency. While in the short term, the errors in the process may be tolerable, in the long-term, the errors accumulate. Why Six Sigma is able to make such a big difference is that it is not just about improving efficiency, but maintaining that state in the long term. For this, a systematic methodology must be followed, that sets clear measurable financial goals at each step, often supervised by a dedicated ‘Champion’.

Three ways of Six Sigma are followed globally – DMAIC, DFSS and Lean Six Sigma.

DMAIC: This methodology has five steps – Define, Measure, Analyze, Improve And Control.

You must clearly define everything before you start – what the current status is (the starting point), and what the project’s endpoints are, in terms of manpower, equipment, and money. e.g. You want to define your project as reducing turnaround time from four days to three for a database-building project.

For this you must measure everything in the current process – what is the number of people on the project, how much time do they spend actually working on it, what is each person’s processivity etc.

Then you analyze that data using statistical and other methods, to find out the causal relationships between components of the process. e.g. You may find that a lot of time goes into training of newbies.

You take steps to improve the process. In this case to speed it up (since your goal is speed) you keep only trained and experienced personnel on the project.

The last step of control is very important, because you must ensure that your goal state, once achieved, does not fall back to the previous state. In the above example, you do this by ensuring that no inadequately-trained person is assigned to the database project, and that those working on it remain motivated to see it through to its logical conclusion.