Customer Satisfaction Measurement Practice in Australia
This study investigates the customer satisfaction measurement practice of Australian managers and analyses the actual and target satisfaction scores reported. The results should be of interest to researchers and to customer satisfaction managers in industry.
This summary is extracted from a journal article by Colin Jevons recently published in the Asia-Pacific Journal of Marketing.
Further information, including a copy of the full paper, is freely available by emailing Colin.Jevons@buseco.monash.edu.au or by clicking on the "info" link at the bottom of this page.
Anecdotal evidence has suggested that a perceived satisfaction score of 50% was unacceptably low for some managers while others considered any score of over 50% acceptable.
A survey of the actual customer satisfaction measurement practices of 503 managers in Australia was therefore undertaken. The mean of the customer satisfaction scores reported by respondents was 83.3%, with an approximately normal distribution. There was a strong relationship between target and actual scores. Whether targets are actually achieved was not strongly related to the magnitude of the target.
There are three main conclusions to be drawn from this work.
- The mean of the satisfaction scores reported was 83.3%, with a standard deviation of 9.63. Some negative skew and kurtosis distort the normal distribution curve, but not to a great extent. It is important to note that a customer satisfaction score of less than 83% is below the average of scores reported in this survey, so the conclusion to be drawn is that organisations with satisfaction levels lower than this should understand that their customers are less satisfied than those of the average of the organisations surveyed here.
- A strong relationship is shown between target and actual scores, but the causality of this is not determined. Whether this is because targets are set with likely actuals in mind, or that actuals tend towards the targets is not known. The conclusion to be drawn is that organisations should be aware of this and respond cautiously. Further research is necessary here.
- Whether targets are actually achieved is not, however, related strongly to the size of the target. The conclusion to be drawn is that it is not necessarily more unrealistic to set a target of 100% than a lower one, although there may be other issues that emerge from setting unachievable targets.
Implications of the Study
There are a number of implications of significance to researchers and practising customer satisfaction and service quality managers.
Understand the Data
It is of vital importance that managers understand clearly the implications of the results of their survey. Managers should understand that a satisfaction figure below 83% is not necessarily, as might be supposed, a good result but is actually worse than the average across those ACSA members who responded to this survey. This is particularly significant in light of the preponderance of ad hoc surveying and measurement systems reported to the researcher during interviews. The importance of professional surveying and analytical practice is emphasised by this result.
Setting Expectations
A high proportion of respondents - the most frequent response was in this category - aim at 100% customer satisfaction. There may well be a number of motivations for this: a zero defect approach, a means of motivating staff, or a promotional idea, for example. Given the popularity of the disconfirmation of expectations paradigm, there is a potential danger in setting customer expectations too high should this be used as a promotional vehicle. If customers are told to expect 100% satisfaction and do not get it, they are likely to be more dissatisfied than if they had not been promised perfect service.
Limitations of the Study
Since the population surveyed was the (then) membership of the Australian Customer Service Association, there is clearly an element of self-selection bias, since a commitment of money and time was involved. While it may well be that the population could be seen as opinion leaders, any generalisation from these results should be done with care.
Additionally, there is the possibility of bias in responses due to self-selection by customers of the organisations surveyed, and because it is likely that a larger proportion of satisfied customers will participate in surveys than exists in the total customer pool.
It is also important to understand the role of customer satisfaction measurement; as McColl-Kennedy and Schneider (2000) have pointed out, the act of measurement affects future interactions with those customers, and that measurement and documentation is not as important as the actual nurturing of relationships with customers.
Further information, including a copy of the full paper by Colin Jevons, is freely available by emailing colin.jevons@ultrafeedback.com or by clicking on the contact links at the bottom of this page.