The term "Cost of quality" seems to be very simple and self-explaining, but the background is more complex. Many stakeholders in various sectors use it incorrectly, and think
that cost of quality is equal to the cost, which comes from producing non-conform parts. Of course this is not true, and cost of quality means much more.
First we can summarize the big three sections of Cost of Quality, which are:
- All investment made to prevent non-conformance, which is the proactive avoidance of failures.
- Assessment and appraisal of products, which means a reactive filtering of non-conform products.
- All costs, that arise for failing to meet customer requirements, and a reactive measure in order to correct issues that has already happened.
The first two sections are together called as "cost of conformance" or "cost of good quality (CoGQ)", while the third section is equal to the "cost of non-conformance" or "cost
of poor quality (CoPQ)".
In order to categorize quality costs, each of them must be analysed based on its origin, intention and aim. As a general formulae: CoQ = CoGQ + CoPQ. The following structure
shows the "Cost of Quality" summary, including examples:
|The quality cost structure
|Cost of Quality
|Cost of Good Quality (CoGQ) / Cost of conformance
|Cost of Poor Quality (CoPQ) / Cost of non-conformance
|Appraisal and assessment costs
|Internal failure costs
|External failure costs
|Quality planning, APQP
|Incoming inspection of purchased materials
|Proactive quality training
|Process audits, Run@Rate
and trial runs
|Additional administration costs
|Quality improvement projects
|Delays and downtime due to quality issues
|Reputation / prestige loss
|Statistical Process Control (SPC)
|Failure analysis and problem solving
|Error proofing, Poka-Yoke solutions
Source: qMindset.com; Philip B. Crosby: Quality Is Free
Increasing CoGQ can decease the whole cost of quality. How can it be? The answer is easy. If you spend more money (and resources) up-front during the planning and design phase,
you can save much more in the serial phase, by decreasing the internal and external failure costs. So the relation between CoGQ and CoPQ is not linear.
Example: you invest $1 million more in the design phase, which results better product-, process design and testing concept, so you may avoid $4 million of failure cost during
the product life-cycle.
Both the CoGQ and the CoPQ curves are exponential. As you invest more in CoGQ, you improve your quality level, however it increases exponentially, you spend more and more to
reduce your failures, and produce only good products. On the other side, your CoGQ indirectly decreases your CoPQ. The two curves have a cross, which is a theoretical break-point, where your cost / quality is
optimized. Over this point (going right on the X axis), your CoGQ is increasing on a higher rate compared to the reduction of your CoPQ. Of course it doesn’t mean you cannot invest in quality over the optimum.
Quite a lot of companies that produce premium products invest in quality over the breakpoint (optimum), and this justifies higher product price. This is one of the reasons, why premium products cost more than
others with lower quality.
Juran is discussing cost of quality from the optimum standpoint (there is an optimal quality level with the lowest quality cost), however Schneiderman goes further with the idea
that is aiming zero defect through continuous improvement. Perhaps going from 3 ppm to 1 ppm may have no high incremental benefit, but brings organizational pride, reputation, spillover into other areas, and
experience in problem solving. In addition, according to Schneiderman: the optimum quality level equals zero defects.
The concept of quality cost (Source: Joseph M. Juran’s Quality Control Handbook; Optimum Quality Costs and Zero Defects: Are
They Contradictory Concepts? by Arthur M. Schneiderman)
Many companies only measure and analyse internal and external failure costs, however the cost of quality analysis should include all quality related costs. There are software
and research companies for this purpose.
Source: qMindset.com; Joseph M. Juran; Arthur M. Schneiderman