Customer satisfaction and investment: Can different operationalizations provide reliable results?

Scientific Essay 2011 16 Pages

Business economics - Business Management, Corporate Governance


Table of Contents

1. Introduction

2. Methodology Selection of data

3. Definition of customer satisfaction

4. Measurement of customer satisfaction

5. Results

6. Conclusions

7. References

1. Introduction

The question, whether customer satisfaction is a reliable indicator for successful financial investment is a rather new field of study. Mainly promoted by Fornell and his associates,[1] the question at first sight, seems to be an empirical question, i.e. a question whether customer satisfaction has anything to add to prediction accuracy of financial performance models. However, this macro-economic view on customer satisfaction is just one of two possible views, the other being the question whether individual investor’s satisfaction provides any kind of clue as to future investment decisions taken by the respective investor. Already, the smooth surface of customer satisfaction models advocated in order to show correlations between investment and satisfaction begins to get wrinkled. Further worry lines emerge, once questions like “is customer satisfaction measured in a consistent fashion?” or “is customer satisfaction unanimously defined?” are asked. This brief research paper will ask exactly these questions and it will do so for the newly emerging field of research linking financial investment and performance with customer satisfaction. Given this agenda, the structure of this paper is straightforward. First, results of a search, gathering papers that deal with customer (Investor) satisfaction and financial performance or financial investment will be reported, then the respective definitions of customer satisfaction will be scrutinized for communalities and differences. Furthermore, differences in the measurement of customer satisfaction will be reported and assessed, and finally, results of the research scrutinized in this paper will be reported before the background of the results gathered so far. A summary of results and the most important conclusions to be drawn from this research will be reported in the final chapter of this paper.

2. Methodology Selection of data

Data reported in subsequent sections is based on a search in different data bases determined by the search terms “customer satisfaction”, “investment”, “investor” and “market”. After a review of the results due to criteria like usability and adequacy, a total of eleven studies remained for further study. Results, reported in subsequent chapters will be mainly based on these eleven studies. In some sense, the results represent a meta-analysis of the respective subject, however, I will not perform a meta-analysis in the proper sense given to the term, e.g. in psychology. The methodology of meta-analysis is designed to give researchers a clearer view of a field upon which a plethora of singular results is scattered: “Meta-analysis allows researchers to arrive at conclusions that are more accurate and more credible than can be presented in any one primary study or in a nonquantitative, narrative review”.[2] However, for a meta-analysis to be able to deliver just what Rosenthal and DiMatteo expect it to deliver, it is necessary that a set of preconditions is met. First of all, it is important to make sure that all studies included in the sample agree with respect to the dependent variable.[3] This in particular is an important demand because quality of a meta-analysis depends on sample-studies being comparable at all. The most important demand with respect to sample–studies, therefore, is that they share the same operationalisation of crucial variables. Studies dealing with customer satisfaction, e.g., should not only have a definition of “customer satisfaction” in common, they should also agree to some extent on the way, customer satisfaction is to be measured (or operationalized). The problem associated with this topic is hotly debated under the headline of “garbage in – garbage out”.[4] Furthermore, meta-analyses aim to deliver a close to comprehensive view of their field of research. Consequently, it is of some importance to identify most if not all of the relevant studies. But: “[e]very meta-analysis has some inherent bias by virtue of the inclusion/exclusion criteria and the methods chosen to review the literature. Not every computer-assisted search will be complete, and not every journal article identified”.[5] Accordingly, every meta-analysis is to some extent incomplete which means that meta-analyses cannot claim to provide a representative picture of the surveyed field. This the more so, because most meta-analysis, as, e.g., Hogan laments with respect to meta-studies that are endemic in the field of personality research, do not check the studies in their sample with respect to definition, operationalisation and measurement of the most crucial variables.[6] This caveat, though it does not finish-off meta-analysis as a tool to bring structure in a scattered field, nevertheless restricts scope and range of meta-analysis to a considerable degree. But this is only the case, if the technique of meta-analysis is used in inductive fashion in some sort of “data speak to me approach”. The value of meta-analyses changes completely, when the method is considered within a deductive framework, which is designed to test a hypothesis. A deductive design to test hypothesis has been suggested by Karl Raimund Popper in his “logic of scientific discovery” which he has called “falsificationism”.[7] In accordance with the methodology set out by Popper subsequent chapter will check the hypothesis that the question whether customer satisfaction has an impact on investment or investment performance cannot be answered due to diverse definitions and measurements taken to delineate and operationalized customer satisfaction.[8]

3. Definition of customer satisfaction

What customer satisfaction is meant to be, for some researchers is so apparent, that it needs no mentioning at all, let alone a definition,[9] while others are convinced that “satisfaction is an overall post-purchase evaluation”[10] and some authors think of satisfaction in general and customer satisfaction in particular as a cognitive process.[11] However, a brief look in the literature shows quite considerable deviations in what customer satisfaction is expected to be. Howarth and Seth define customer satisfaction as the cognitive state of being adequately or inadequately rewarded for the sacrifices one has made.[12] In a widely used model, customer satisfaction is integrated in the confirmation/disconfirmation paradigm.[13] According to this model, customer satisfaction is defined as the difference between customers’ expectations of a particular product or service and the actual performance of the respective product.[14] Although being a definition that is widely shared, if not to the degree suggested by Fornell,[15] some modification even to this definition exist. E.g., Locke modified the model by adding a third variable, i.e., the importance a service or product has for a particular customer.[16] Consequently, customer satisfaction is the difference between expectation and performance multiplied by the importance of the product. Depending on whichever definition is deployed in a study, results will vary. Unfortunately, this is not the entire list of competing definitions. Definitions differ with respect to affective connotations, i.e. some researchers define satisfaction explicitly as something that appears if “the experience was at least as good as it was supposed to be”.[17] Others conceptualize satisfaction as a discrete or continuous variable with the extremes satisfied and dissatisfied.[18] Hence, satisfaction does not equal satisfaction. Different studies include different concepts of “satisfaction”. Nevertheless, Fornell[19] was right in claiming that most definitions of satisfaction refer to an assessment or an evaluation that follows a purchase and compares the performance whatever has been purchased to performance expectations formed pre-purchase. But, as Kanning and Bergmann[20] outline, not even the widely shared minimum agreement about satisfaction being the result of some kind of comparison between performance and expectation provides solid ground, because (1) expectation is a rather vague concept, (2) expectations may be exceeded and it is not clear how this would figure in the concept of “satisfaction”, (3) the assumption that expectations and performance have the same importance. None of the aforementioned problems is of some relevance to the authors of the eleven studies surveyed in the meta-analysis performed in this paper. While those authors using the American Customer Satisfaction Index (ACSI) simply stated, that what readers must know about customer satisfaction in general and the ACSI in particular has been said by Fornell[21], the remaining authors, with the notable exception of Kanning and Bergmann, do not even bother about a definition for customer satisfaction.[22] It seems all to clear to these authors, what customer satisfaction is meant to be. However, agreement about customer satisfaction’s content is brittle. A look at the measurement of customer satisfaction deployed in the surveyed studies will show this.


[1] E.g., Anderson, Fornell & Mazvancheryl (2004); Fornell (2007); Fornell et al. (2006).

[2] Rosenthal & DiMatteo (2001), p.61.

[3] Borenstein et al. (2009), pp.26-28.

[4] Hunt (1997), p.18.

[5] Rosenthal & DiMatteo (2001), p.66.

[6] Hogan (2005), p.334.

[7] Popper (1959), chapter IV.

[8] The way meta-analysis is used in this paper makes it avoidable to invest a lot of time and nerves in the attempt to make correlation coefficients and statistical procedures used in different studies compatible. Accordingly, no effect-size statistic will be calculated in the course of this paper. “The effect size statistic produces a statistical standardization of the study findings such that the resulting numerical values are interpretable in a consistent fashion across all the variables and measures involved. … The key of meta-analysis, therefore, is defining an effect size statistic capable of representing the quantitative findings of a set of research studies in a standardized form that permits meaningful numerical comparison and analysis across the studies”; Lipsey & Wilson (2000), pp.4-5. This quote makes it all to clear that meta-analysis rest on the assumption that results gathered in different studies can be standardized. This assumption is only feasible if studies show at least a minimum amount of “common variance”. As research has shown time and time again and as especially Hunt (1997) has most sophisticatedly pointed-out, this usually is not the case.

[9] E.g., Ding et al. (2008).

[10] Fornell (1996), p. 11.

[11] Giering (2003), p.20.

[12] Howarth & Seth (1969), p.145.

[13] Davis & Heineke (1998).

[14] Kanning & Bergmann (2009), p.379.

[15] Fornell (1996), p.11.

[16] Locke (1976).

[17] Hunt (1977), p.72.

[18] Tse & Wilton (1988), p.205.

[19] Fornell (1996), p.11.

[20] Kanning & Bergmann (2009), p.379.

[21] Fornell (1996).

[22] Kanning & Bergmann (2009).


ISBN (eBook)
ISBN (Book)
File size
452 KB
Catalog Number




Title: Customer satisfaction and investment: Can different operationalizations provide reliable results?