How to Implement a Solution with Six Sigma
After you use the Six Sigma methodology to identify the root cause of the problem, you need to select and implement a solution. Chances are the problem has many solutions. So a decision matrix can help you decide which solution to pursue.
A decision matrix, often called a Pugh matrix, evaluates alternative solutions by comparing each along a series of objectives. Often, the current method serves as a baseline.
Say your team comes up with three solutions (A, B, and C) to a problem. Follow these steps to construct a decision matrix:
- Select the solutions you want to evaluate. Find out what improvement you can expect from each solution. The solutions are listed in the top row of the matrix in the table below.
- Decide what the criteria for evaluation are. Choose the most important criteria that relate to the issues that made you decide to focus your efforts on the project. The criteria are listed in the first column of the matrix in the table below.
- Score each solution against each of the criteria.
Always give the baseline a score of 0. Typically, a 3-point scale is great for this, but if more differentiation among the solutions is desirable, you can use a 5-point scale. In a 5-point scale, use the following numbers:
+2 Much better than baseline +1 Better than baseline 0 Equal to baseline –1 Worse than baseline –2 Much worse than baseline
- Rate the importance of the criteria if necessary. If one or more of the decision criteria are considered more critical than the others, you can assign a weight to each. Then multiply the score from Step 3 by the criteria ranking, giving it a final score.
- Sum the assigned scores. Add the scores for each solution along the criteria to get a net score.
- Choose your solution. In most cases, you want to choose the solution with the highest score. (For example, in the table below solution B is best.) If all alternatives score less than zero, the baseline (current) process is considered the best option. You may want to consider other solutions if the current batch doesn’t offer improvements worth the cost of implementation.
Keep in mind that rating the criteria, as described in Step 4, may have a great effect on your solution decision. For example, if you rate the importance of criteria 1 higher than the others, solution C may well end up being the preferred solution.