All of the qualities listed and implied in the previous sections have a direct bearing on the product's Quality; and by the definition given herein, all of the aspects of Quality have measurable or assessable goals, which if met, would yield a product of 'optimum' Quality. It's important to note that the specific goals for some qualities may change for various reasons over time. Consequently, the product development team needs to be aware of any changes in the goals (and associated tolerances) so that the Quality of the product can be monitored and maintained.

The three categories of Quality must be defined as soon as possible in the development process; and... it must be done for each specific product and process since no two products or processes are identical. Every product quality, which is important to the Quality of the product, must be identified, mapped into a measurable or assessable goal with a commensurate tolerance so that the control of product variability is possible. This allows one not only to talk-the-Quality-talk, but also to walk-the-Quality-walk. This process is commensurate with Taguchi's concept of "Off-line Quality Control"; i.e. designing for robustness against variability in production and the user environment.

Q1 and Q2 need to be unearthed and set before the technology is selected. Then each respective Quality goal and associated tolerance need to be set. These become part of the Marketing Requirements Document (MRS), which is used eventually to validate that the correct product is designed by testing the product to the claims of the MRD. Quality Function Deployment is one of several good techniques to use for this task (reference 3). When the technology is selected, the Q3 qualities need to be selected appropriately to yield the desired technical and production Quality. These become part of the Product Requirements Specification (PRS), which is used to verify that the product is designed correctly by testing the product to the claims of the PRS.

One can use equation [1] or [2] to keep track of the Variability if the qualities are small in number. In some design situations, there are so many qualities that all can not be accounted and tested for in a simple manner. In complex cases Statistical Design Of Experiments (SDE or DOE; see reference 2) is an analytic methodology that can be used to statistically determine a much smaller set of qualities, which can be tested quickly to track effects on product Quality. Many semiconductor companies use this technique when testing new ICs, (because of a large number of variables). But for a great many products the variables are manageable by a much simpler method as follows.