Measuring quality

​
​It is a fundamental tenet of management that to control something you first need to measure it. This presents a very real problem for quality management. As we have previously discussed, there are different understandings of the term 'quality', so there is no agreement about what should be measured.

There are three types of measure that might be used, each based on a different definition of quality:

Financial measures. These are based on the concept of the costs of quality (prevention, appraisal and failure). It is argued that measuring quality in the common language of finance makes it accessible to all managers. This not only puts quality firmly on the broader business agenda within an organisation, but enables quality decisions to be taken objectively by the weighing of costs and benefits in financial terms.  

However, there are problems with this approach. There is no commonly agreed basis for measuring the costs of quality, and in any case any accounting system is based on a series of managerial judgements (how to allocate fixed costs and overheads, what depreciation method to use, and so on).

Operations-based measures. These tend to be based on the ‘conformance to specification’ approach to defining quality. It tends to be easier to use such measures in manufacturing operations, in which the products and the processes used to produce them are quite distinct.  
​
The physical attributes of goods can usually be measured with precision, and aspects of their performance can be tested. The performance of the production process can also be measured by, for example, determining the level of defect or scrap rates produced. Various aspects of service operations can still be measured for conformance to specification, bur some of them can be controversial, for example judging the quality of a health service by the length of its waiting lists.

Where large quantities are involved, it may be appropriate to use statistical techniques such as acceptance sampling and statistical process control charts. Acceptance sampling is used to decide whether to accept or reject a batch of goods on the basis of checking a few items from the batch; statistical process control charts are used to monitor the performance of a process by checking its output at regular intervals.

The results can be used not only to confirm whether a current output is within the permitted range but also to give warning of likely problems and hence the need for preventative action.

Customer-based measures. These are measures that stem from the user-based or value-based approach to quality. They are typically based on attempts to gauge the degree of customers' satisfaction. This may be done by using the feedback forms that are so prevalent in hotels, restaurants, airlines, and so on.

Alternatively, customers may be interviewed by organisational staff or independent market researchers; or focus groups and the like may be used. Customer-based measures tend to be more prevalent in service organisations.

Given the intangible nature of most services, this is probably necessary and desirable. This approach requires quality to be defined very much from the customers’ viewpoint, and so is bound to be based on their perceptions; but such an approach may lead solely to an analysis of failure, with little or no attention given to emulating successful situations. This approach can be applied to internal and external customers.

​Whatever measures are used, it is still necessary to know why the measurement is being carried out, and what will be done with the results. Measures in themselves are not particularly useful. They are useful only when used comparatively. This requires some kind of standard or target against which to compare current performance.