Cost of Quality

If we were to go out and survey your average customer on the street and ask them “What level of quality would you like in the next product you buy?” Most of them would gaze back from under a raised eyebrow and say “I want the best quality.” Or something along those lines. Similarly, they would also probably say that they want a low cost as well. Every supply chain decision is made in this framework with the tensions of cost and quality pulling often in different directions (for those of you following along in your books, one could also add in delivery and service level too!).

The Six Sigma approach to Cost of Quality is to divide the cost into internal and external costs, and then divide the internal costs further into cost of Good Quality and then the cost of Poor Quality. This is fundamental to the Six Sigma approach because before any major change project can proceed, calculating the return is a key decision factor. In practice though, decisions are rarely obviously clear, analyses rarely this accurate, and circumstances are rarely this static. In modern manufacturing environments and supply chains, change is the norm and learning to live with it, while continuing to push for improvement is a key to survival. Having a firm foundation in principles such as the cost of quality and adhering to a process is critical, but don’t forget intuition is often a good guide too.

So, what should we do to decide if pursuing quality improvements are worth the resources and how should we run a cost reduction program?

The best way that I have found to make these kinds of decisions and to run a quality program is to define up front the analysis that we will do on each major change and then generally follow a standard process for making decisions on how to prioritize the team’s efforts and then methodically work our way down the list of improvements. Periodically we need to reprioritize in a monthly or quarterly quality meeting with all key stakeholders participating.

As a normal course of production, there should be IQC inspection and IPQC inspections with data collected and reported for each inspection. Further, we need to understand the rework costs for each type of defect. Generally we will see defect distribution arranged according to some sort of a Pareto law and it will be easy to determine which defects are the biggest pain points. As a normal practice the process quality leaders should be tabulating this data, pursuing root causes for each defect, looking for patterns, proposing improvement actions, and reporting status. This is all straightforward, and should involve little decision-making by executives. Once yields are “high enough,” then quality tends to get even less attention until something goes wrong—and it’s really only a matter of time, by the way.

The difficulty arises when the nervous, over-caffeinated, sleep-deprived quality engineer says in the big meeting, in a voice two octaves higher than normal, “we need new equipment…” The long silence that follows is characteristic of an organization that has not fully come to grips with how to approach these decisions. Think of the poor quality engineer who has been dreading bringing this up and had no clear idea of what the expectations were around asking for replacements or upgrades. Also, think of the unfortunate surprise experienced by the other stakeholders as they suddenly find out that there are problems that are going to cost money and take a bite out of their P&L that quarter. Think of the Production engineers who are suddenly very interested in the ceiling for some reason as their eyes all turn upward to study the ceiling tiles.

The best way to avoid this situation is to plan for the eventuality that not everything in the world goes exactly as planned and put a simple, lightweight expectation in place for how you want to manage quality. Make a Plan for Good Quality up front, and do the homework. Collect baseline data, do process capability analysis and calculate the Cpk’s up front so you’ll have an idea of your starting point and whether or not the process is capable. Assuming it is, then most of the problems should be minor and solvable through smart thinking, good communication, and a little bit of money here and there.

The return on this investment will certainly be more than the cost of defects piling up and unhappy customers demanding replacements and refunds. And your team will be much more happy too!