At all costs?

The concept of the ‘cost of quality’ is not new. But is it time to revise how this is assessed, asks Dr Ron Basu.

Juran first discussed the cost of quality analysis as far back as 1951. However, it is Feigenbaum who should be credited with the definition of the cost of quality (COQ), when he identified the four cost categories in 1956. These can be classified as prevention costs, appraisal costs, internal failure costs and external failure costs. Both concept and categories have been followed in basically the same format ever since.

Prevention costs and appraisal costs are often defined using one of three terms: as the cost of control; the cost of conformance; or the cost of good quality.

Regardless of the label used, this refers to the outlay of setting up and managing a quality management team with clearly defined processes. Similarly, internal failure costs and external failure costs are also combined to be known as the cost of failure, the cost of non-conformance or the cost of poor quality. These are the expenses of defects and reworks arising from poor-quality management.

Thus, the basic theory is quite simple; however, the challenge lies in defining and measuring each of the sub-components leading to the four major cost categories. As they say, the devil is in the detail. It is a bit like having a baby – easy and pleasing to conceive, but rather more painful to deliver!

Traditonal vs modern views
The traditional view is that the cost of control increases and the cost of failure decreases until it reaches a point beyond which the total cost of quality increases.

In other words, from this point, the cost of improving quality becomes larger than the derived benefits. This is sometimes called the ‘optimum point’ of quality effort (see Figure 1).

However, with the application of Total Quality Management (TQM) and Six Sigma processes, the traditional view outlined above becomes challenged. Some precision manufacturing operations (and even service operations such as air transport) can accept zero-defect or Six Sigma standards.

Furthermore, the traditional model implies that the cost of control is high in proportion to the cost of failure. Increased quality is not achieved by more inspectors, but instead by a culture of quality assurance underpinned by initial training. As time progresses they may incur some charges for additional training for improved processes; however, this will not be at the same rate as for the early stage of a quality programme.

If we incorporate corrections from the total quality culture then the time-honoured cost of quality model would change to the representation as shown in Figure 2.

It is true that there may not be enough data to validate the optimum point of the total quality cost. However, the significant savings generated by project-based Six Sigma or Lean Sigma programmes supports the fundamental argument. This is quite simply that a diminishing total cost of quality is achieved by following a sustainable continuous quality improvement programme.

Coq in the construction industry  
Now we can reason that activities related to cost of quality are showing favourable results in both the manufacturing and service industries, thanks to holistic quality programmes.

However, the picture is very different in the field of project management. A most significant gap in the project management body of knowledge is the cost of quality. The application of Six Sigma in major projects is also limited. Nonetheless, one project sector in which the potential of COQ is beginning to be recognised is construction.

While we appreciate the role of quality in construction projects, we do need to understand much more about how we can define and measure the dimensions of the cost of quality.

In particular, it is vital to gauge the external costs of failure due to poor-quality delivery during design and construction. In turn, these external expenses then affect the operational performance and outlay of assets post-handover.

Anecdotal evidence suggests that construction project insurance accounts for 1.5 per cent of project costs. If we conclude that this is caused by a failure in quality then the figure we are looking at amounts to as much as 1.5 per cent of £120bn (ie £1.8bn) annually in the UK alone – an astronomical sum.

More reassuringly, there are further nuggets of wisdom to be gleaned that will help us to ameliorate this sort of situation. In order to find them, we need to define and measure by empirical research the key metrics of COQ as applied to construction projects.

The coq metrics  
As a start, let us attempt to apply Feigenbaum’s four categories of COQ with some typical examples of the metrics of each category for construction projects. We would develop Table 1 as an outline, as shown opposite.

The metrics in each category are shown as examples only and these are extended depending on the nature of the projects and quality specifications. Cost of control is usually easily measurable at the early stage of the project life cycle. Major construction projects have already started measuring and monitoring internal failure costs during the design and construction phases of the project. But a key area that remains neglected is the assessment of external failure costs, which occurs after the project handover. It is a crucial omission since, arguably, this is the aspect needing most attention.

The costs after the handover are likely to provide clients, designers and contractors with data in order to focus their efforts to improve the biggest contributors to the cost of quality.

The way forward
In order to make the ‘painful’ delivery process a success, here are some fundamental principles of quality and performance management. These include:

• Each metric should relate clearly to one of the specific four categories.
• The metrics must follow the rigour in objectives, rigour in measurement and rigour in use.
• Each metric is unique, without any duplication or overlapping in the area of measurement.
• Each metric should be clearly described with a definition, formula (if applicable), worked example and purpose.
• The metrics should be tested in a pilot study before they are implemented.
• Metrics in each category should lead to root-cause analysis by identifying whether it is design-related, process-related or people-related.  

---

Dr Ron Basu is director of Performance Excellence and a visiting fellow at Henley Business School