By Chew Jian Chieh
What is the usual way most people go about solving problems? Most people and organizations consciously or unconsciously use this method, as illustrated in Table 1 below.
| Table 1: Normal Method of Problem Solving |
Step | Example |
1. Understand what is to be
improved and set a goal | I am too fat.
I want to reduce my weight. |
| 2. Measure their current state | I am currently 90 kg. Ideally,
I should be about 70kg. |
3. Apply conventional wisdom
or gut theory | If I exercise more and eat less,
I should lose weight. |
| 4. Take action | Exercise more and eat less. |
5. Measure to verify improvement
has taken place | I lost 5kg. |
This is not a bad method, provided what one thinks is causing the problem is really causing the problem. In this case, if a person is fat simply because they do not exercise enough and eat too much, then by exercising and eating less, they should weigh less. And if they do lose weight after taking such action, then the theory is validated. People solve a fair number of problems in this manner – using conventional wisdom and gut theories that also happen to be correct. In those cases, there is little need for Six Sigma – it is just a waste of time. Just do the above.
How Six Sigma Problem Solving Is Different
How is the Six Sigma problem-solving methodology different? Actually it is really not so different from how people normally go about solving day-to-day problems, except in Six Sigma, nobody knows what is really causing the problem at the beginning of the project. And because all attempts to solve the problem in the past have failed, largely because conventional wisdom and gut theories were wrong about the cause of that problem, people conclude that the problem cannot be solved.
These types of problems are really the best candidates for Six Sigma. The Six Sigma DMAIC methodology differs from conventional problem solving in one significant way. There is a requirement for proof of cause and effect before improvement action is taken. Proof is required because resources for improvement actions are limited in most organizations. Those limits preclude being able to implement improvement actions based on 100 hunches hoping that one hits the mark. Thus, discovering root causes is at the core of the methodology.
Here are the steps in the DMAIC process:
- Define phase: Understand what process is to be improved and set a goal.
- Measure phase: Measure the current state.
- Analyze phase: a) Develop cause-and-effect theories of what may be causing the problem; b) Search for the real causes of the problem and scientifically prove the cause-and-effect linkage
- Improve phase: Take action.
- Control phase: a) Measure to verify improvement has taken place; b) Take actions to sustain the gains.
Using a More Mathematical Language
The above steps can be phrase in another way – using more mathematical language (Table 2). (This kind of mathematical language should not put anyone off. If it is a concern initially, a person just needs to remember than whenever a Y shows up in any sentence, just replace it with word "effect," or the phrase "outcome performance measure." And whenever an X shows up , just replace it with the word "cause.")
| Table 2: DMAIC in Mathematical Terms |
Steps | Phase | Questions |
| 1. Understand what process is to be improved and set a goal. | Define | > What is the Y or the outcome measure? |
| 2. Measure the current state. | Measure | > What is Y's current performance? |
3. Develop cause-and-effect theories of what may be causing the problem.
4. Search for the real causes of the problem and scientifically prove the cause-and-effect linkage. | Analyze | > What are the potential Xs or causes?
> What may be causing the problem?
> What are the real Xs or causes?
> What is really causing the problem?
|
| 5. Take action. | Improve | > How can the understanding of the real causes of the problem be exploited to eliminate or reduce the size of the problem?
> How can this Y = f(X) understanding be exploited? |
6. Measure to verify improvement has taken place.
7. Take actions to sustain the gains. | Control | > Did Y really improve?
> How can the Xs be controlled so the gains in Y remain? |
The key assumption in Six Sigma is this: If the true causes of any problem can discovered, then by controlling or removing the causes, the problem can be reduced or removed. Now is that not just common sense?A Series of Common Sense Questions
In summary, Six Sigma DMAIC methodology is really just a series of common sense questions that one asks in order to solve any problem and eventually sustain the gains that come from solving the problem.
- Define: What is the Y that is not doing well?
- Measure: What is Y's current performance?
- Analyze: What are the potential Xs? What are the real Xs?
- Improve: How can the real Xs be controlled or eliminated?
- Control: How can the Xs continue to be controlled to sustain the gains in Y?
Six Sigma's DMAIC methodology is nothing but a search for the real causes of problems. With this understanding, what remains for those learning Six Sigma are the various tools and techniques used to answer these questions.
About the Author: Chew Jian Chieh is a senior consultant with Valeocon Management Consulting and supports clients in China and across Asia He has extensive experience in implementing process and organization improvements for various industries. He specializes in Lean Six Sigma, Strategy Development/Deployment and Change Management. Chew JC is a Singapore national. He can be reached at jian-chieh.chew@valeocon.com.
