Get the book “Six Sigma for Everyone” by George Eckes, John Wiley and Sons 2003.  It does a very nice job of outlining what to do and provides some really good tools.  Seems like a good book for green belts.

To extend beyond the aforementioned book, let us come to understand the importance of properly “mating” Six Sigma tools with product and process data.  Analogously speaking, think of tools like guns and data like bullets.  If you try to put a .22 caliber bullet into a .38 caliber gun, what is likely to happen?  Certainly, you will not get the anticipated result.  So it goes with statistics and data.  Here is a simple guideline that I have taught for many years:

Step 1: Explain the decision that must be made.
Step 2: Detail the information that is required to make that decision.
Step 3: Specify how you must “see” the outcome of Step 2 (i.e., chart, table, etc).
Step 4: Decide on the type of analytical tool that best drives the outcome of Step 3.
Step 4: Establish the type of data must be gathered to feed the tool specified in Step 4.
Step 6: Identify where the required data must be collected.
Step 7: Design a sampling plan that will properly organize the collection of data.

Please understand that the given order is only for the purposes of planning.  To execute the plan, one must reverse this process – Step 6, … , Step 1.  Far too often, practitioners rush out and simply do things in reverse.  Remember, we must use “reverse planning” and not “reverse execution.”  This simple little motto will help to ensure that all the “i”s are dotted and “t”s are crossed  – before any action is taken.  In the form of questions, the process can be summarized as follows:

What decision must be made?
What do you need to know to make that decision?
What form of visual will provide the information?
What kind of tool will generate the visual?
What kind of data is need for the tool?
Where can the data be found?
What is the plan to get the data?