SPC and the Cost of Quality


The Cost of Quality has been defined as the sum of 4 separate costs: the cost of prevention, the cost of appraisal, the cost of internal failure and the cost of external failure. You can use the Statit SPC tools to understand and adjust these costs.

Cost of prevention is the cost of activities focused on preventing poor quality in the product or service. These costs can be things such as the design process, process capability studies, quality meetings and planning and a host of other activities having to do with preventing failures.

Costs of appraisal are those costs associated with measuring or evaluating conformance to your standards, specifications and quality measures. These include things such as inspections (incoming material and product testing); product, process and service audits; calibrations, as well as the cost of the equipment and materials to appraise the product or service.

Costs of internal failure are those costs associated with finding a failure in-house before it reaches the customer. Rework and scrap are prime examples of these costs as well any other activities associated with a failure found before delivery to the customer.

Costs of external failure are typically viewed as the most expensive and may include some intangible costs. Tangible costs could include processing customer complaints, warranty claims, recalls and returns. Intangibles could include things such as lost opportunities and diminished reputation.

SPC tools are effective in ascertaining the costs of internal and external failures. A weighted Pareto or Paynter chart is a good example of such a tool. For example, the following Paynter chart reflects number of defects in certain categories over the last 6 months.

There is some good information here but we might wonder what it would look like if we assigned a cost to each category. This cost might be the cost to rework or scrap.

This chart gives us much better information on where we are spending the most in relation to defects produced. We see now that when we look at the costs, LS0001 defect code may be the most important defect code to work on.

If we look at a p chart for the LS0001 Defect Code, we see that the process has definitely shifted over the last few months.

Java is not enabled in browser, data tips cannot work for this graph.
(mouse over data points for more details)

With these data, the i chart also shows an issue with the cost of defects. This is not always the case with an i chart.

Java is not enabled in browser, data tips cannot work for this graph.
(mouse over data points for more details)

If the codes in this data were all failure codes, whether external or internal, how could we have avoided this situation? Let’s take a look at a p chart from earlier in the year.

Java is not enabled in browser, data tips cannot work for this graph.
(mouse over data points for more details)

Obviously, we could have prevented much of the failure cost by making a change to the process at this early date. If we can investigate an assignable cause as soon as we detect it and implement a process change, we prevent future failures. And in your process, I would hope that you would be looking at a data period shorter than 1 month so that you can respond in a more timely manner.

We would certainly need to evaluate the cost of prevention (as in changing the process) in relation to the costs of failures. If we are NASA, cost of failure is very high, so our cost of prevention can be significant. In any regard, we would need to weigh the tangible costs of external failure as well as the intangibles, the costs of internal failures and the cost of appraisal against the cost of prevention. One note on appraisal: 100% inspection does not necessarily mean 100% of failures captured.

You can further explore these charts, as well as other quality reporting, on our live demo site, live.statit.com.

If you would like additional information, please send email to statit.support@acs-inc.com.