A Study Of Two Customer Retention Measures: The American Customer Satisfaction Index And The Conversion Model Nic S Terblanche, Department of Business Management, University of Stellenbosch Jannie Hofmeyr, Customer Equity Company, Cape Town Abstract Customer retention has become a major concern for many businesses. Various means to measure and predict loyalty and commitment have been developed to attend to this need. One of management s major challenges is to utilise a model suitable to explain and predict customer retention for a particular company or brand. This paper compares two prominent measures, namely the American Customer Satisfaction Index (ACSI) and the Conversion Model (CM). The respondents were 2000 consumers drawn randomly from the seven major metropolitan areas of South Africa. Respondents were required to respond to items from the ACSI and CM scales in relation to the fast food and motor car industries. The results produced by the ACSI model to predict customer loyalty indicate a weak relationship between customer loyalty and customer satisfaction. These findings are in line with the views of various authors who state that customer satisfaction cannot be a sole predictor of customer loyalty. The CM results differ considerably from those of the ACSI model. Explanations, which are useful to theoretical and managerial perspectives, are presented for these differences. Keywords: customer retention, American Customer Satisfaction Index, Conversion Model Background Every company knows that it costs far less to hold on to a customer than to acquire a new one. That is why customer retention has become the Holy Grail in industries from airlines to wireless (Coyles and Gokey, 2005). Customer retention has become a major area of consideration for businesses as it has been recognised as an important contributor to profitability over the long term (Anderson and Sullivan, 1993; Fornell, 1992; Iniesta and Sánchez, 2002; Heskett, Sasser and Schlesinger, 1997; Heskett, 2002; Dick and Basu, 1994; Anderson, Fornell and Lehmann, 1994). Three measures of customer retention, namely satisfaction, loyalty and commitment, have received wide coverage in academic journals the past fifteen years (Oliver, 1999; Auh and Johnson, 2005; Wetzels, de Ruyter and van Birgelen, 1998). It is especially the relationship between satisfaction and loyalty, and between satisfaction and commitment that have received a great deal of prominence in the academic literature (Oliver, 1999; Clerfeuille and Poubanne, 2003). However, not all satisfaction (even at a high degree) translates into loyalty (Mittal and Lassar, 1998). Miranda, Kónya and Havrila (2005) established that shoppers satisfaction levels are not the only key to store loyalty indeed they found no evidence in their study that shoppers overall satisfaction was by itself a significant influence on continued patronage. Andreassen and Lindestad (1998) also found that, contrary to their expectations, customer satisfaction has no impact on customer loyalty. Some argue that trust is a stronger emotion than satisfaction and it may therefore better predict retention (Hart and Johnson, 1999). Customer retention has also often been conceptualised and operationalised as a dimension of the loyalty construct (Boulding, Kolra, Staelin and Zeithaml, 1993; Zeithaml, Berry and Parasuraman, 1996). The dilemma for managers is that even when their customers say they re satisfied they still switch to other suppliers or brands. The truth is that satisfied customers aren t necessarily loyal customers. ANZMAC 2005 Conference: Relationship Marketing (Consumer) 71
Loyalty requires a commitment from the customer that mere satisfaction cannot bring. The implications are that customer satisfaction measures are inadequate on their own and need supplementing by a measure of loyalty and it means that one cannot focus on those elements of quality creating satisfaction because they don t encourage loyalty. Purpose The phenomenon of customer retention has a degree of fuzziness since it embodies a theoretical construct which cannot be observed directly (Gerpott, Rams and Schindler, 2001). The ways in which academics and practitioners define customer retention, whether conceptually on empirically, show substantial variance. The approaches in which customer retention is distinguished from related constructs such as customer loyalty, customer satisfaction and customer commitment, also have great differences. The primary purpose of this paper is to report on the empirical findings of a study undertaken to compare customer retention figures as determined by the ACSI and the CM. Both these models are used globally to measure and predict customer retention. The outcomes of the two models, namely customer loyalty in the case of the ACSI and customer commitment in the CM, are the measures used for comparing the customer retention determination of the two models. A secondary purpose of the paper is to explain the operationalisation and the inherent differences between the two models. For the purposes of this study, customer satisfaction is a direct determining factor for both customer loyalty and customer commitment. Customer retention, in turn, is regarded as outcome of the respective models. The outcomes of the two models studied, namely customer loyalty in the ACSI and customer commitment in the CM are regarded as similar for purposes of comparing their abilities to predict expected future behaviour. The Two Measures Employed The two measures employed in this study are the ACSI and the CM. A brief explanation of the two measures is provided to highlight the major differences of the two measures. American Customer Satisfaction Index A primary objective of ACSI is to estimate the effect of ACSI on customer loyalty; customer loyalty is the ultimate dependent variable in the ACSI model because of its value as a proxy for customer retention and its effect on profitability (American Customer Satisfaction Index, 2001; Fornell, Johnson, Anderson, Cha and Bryant, 1996). In the ACSI model three antecedents drive satisfaction. The three antecedents are perceived quality, perceived value and customer expectations. Perceived quality can be split into perceived product quality and perceived service quality. The latter is used when the so-called expanded ACSI model is applied. The expanded model was applied in this study. All the latent variables in the ACSI model is measured by three manifest variables (items), except for perceived value and customer complaints which are measured by two items and one item respectively. The Conversion Model The CM establishes a consumer s level of commitment to a brand to make predictions about future loyalty. The CM uses four dimensions to measure commitment. The first dimension is satisfaction with the brand. However, satisfaction correlates poorly with future behaviour, as ANZMAC 2005 Conference: Relationship Marketing (Consumer) 72
evidenced by many researchers (Mittal and Lassar, 1998; Miranda, Kónya and Havrila, 2005; Andreassen and Lindestad, 1998). Understanding satisfaction does not help us to fully understand why consumers do what they do. Satisfaction is, however, an essential component of understanding the relationship between consumers and brands. The second dimension is a customer s perception of the alternatives. The evaluation of a brand does not happen in isolation. It is a comparative measure against the competition. This also means that a high customer satisfaction score does not necessarily mean that the relationship with the customer is secure. The third dimension is the importance of brand choice to a consumer. If brand choice doesn t matter, it is difficult to achieve commitment. The product category, as well as brand choice, has to be something relatively important in the consumer s life for commitment to be possible. The more that brand choice matters, the more likely it is that the consumer will take time and trouble to make final decision about which brand to choose. The fourth dimension is the consumer s degree of ambivalence. The more ambivalent the consumer is about which brand to choose, the more likely it is that the final brand choice will be delayed until the last possible moment. For consumers such as these, point-of-purchase stimuli become critically important, as they will often only make their final choice at the shop shelf. Measures Methodology and Empirical Study The scales used for the ACSI model, are those set out in the ACSI methodology report (American Customer Satisfaction Index, 2001). The dimensions measured were customer expectations, perceived product quality, perceived service quality, perceived value, customer satisfaction, customer complaints and customer loyalty. Most of the dimensions (except for customer complaints and perceived value) were measured with three items. The items were measured on 10 point Likert scales, ranging from not very high/well/often/almost/never/very poor/very dissatisfied/ etc (1) to very high/well/often/almost always/ very good/very satisfied/ etc (10). Customer complaints were measured by whether a respondent complained in the past 6 months (Yes/No) and how well the complaint was handled on a 10 point Likert scale ranging from handled very poorly (1) to handled very well (10). Perceived value was measured with two items on a 10 point Likert scale. The CM employs the following four dimensions to measure a consumer s commitment to a particular brand (Rice and Bennett, 1998; Hofmeyr and Rice, 2001): Need satisfaction which refers to the extent to which consumers needs are satisfied by a particular brand. Involvement in the category which express the importance of the brand choice in the category to the consumer. Attitude to alternatives which indicate the levels of commitment that the consumer has to the brands currently used. Intensity of ambivalence - which refers to the degree to which the consumer is pulled in different directions by the brands on offer. Two of the four dimensions were measured with one item each per dimension on 7 point Likert scales and one on a 10 point Likert scale. The intensity of ambivalence was measured on a 3 point scale. ANZMAC 2005 Conference: Relationship Marketing (Consumer) 73
Data collection A structured questionnaire was used and it was personally administered. The interviews were conducted with respondents during the period 13 to 27 June 2005. All interviews were conducted using random suburb sampling to obtain a representative sample. Suburbs were drawn from the 2001 South African census and five interviews were conducted within each selected suburb to ensure that a demographically representative sample was achieved across suburbs and areas. The total sample for this study was 2000 interviews. Other characteristics of the sample and sampling procedure are as follows: the interviews were conducted in the seven major metropolitan areas of South Africa the sample is in the same proportions of the South African population in terms of both ethnic group and gender to be representative respondents are interviewed in-home respondents were required to be 18 years and older to qualify for the interview. Results The purpose was to use ACSI to predict customer loyalty, where customer loyalty is based on customer satisfaction and customer satisfaction is the outcome of customer expectations, overall product quality and product value. As can be seen from Figure 1 and Table 1 (the results i.r.o. the fast food retailers), 45% (from R 2 ) of the variance in customer loyalty is explained by customer satisfaction. The balance, 55%, can be ascribed to measurement errors in customer satisfaction and loyalty, together with the influence of other unknown factors. The composite reliability, which represent a measure of how well a latent variable is measured by its manifest variables (items), indicated a lower value of 0,549 for customer loyalty than for any of the other latent variables. This could be seen as a contributing factor for the relatively low R 2 of loyalty. Figure 2 and Table 2 contain the results of the analysis for motor car manufacturers. In this case, only 36,8% of the variance in customer loyalty is explained by customer satisfaction. The composite reliability is less than that for the fast food retailers, again indicating that the items used to measure customer loyalty is not very reliable. The small percentage of loyalty that is explained by satisfaction could be ascribed to the fact that loyalty is not measured properly. Box plots were done to gain a better understanding of the items measuring loyalty. The two items used for the measurement of loyalty produced contradictory findings. Bootstrapping was applied to estimate the confidence intervals of the R-square for customer loyalty. Five hundred samples of 500 respondents each were used. The observed values for different confidence levels are set out in Tables 1 and 2. Partial least squares (PLS) was used to estimate the relationships in the ACSI model. PLS is an iterative estimation procedure that integrates aspects of principal-components analysis with multiple regression and is well suited to predict the relationships in the ACSI model (Gustaffson and Johnson, 1997; Steenkamp and van Trijp, 1996; Wold, 1982). The customer complaint dimension was not included in estimation calculations because very few respondents indicated that they have complained within the required period. The CM classifies consumers in terms of their level of commitment to a brand. The relationship that consumers have with every brand in the marketplace is identified in this process. This enables one to categorise users and non-users of a particular brand in eight segments. Figures 3 and 4 illustrate the commitment categories for the fast food and motor car industries. The next step was to calculate the CM equity score for each brand. The equity ANZMAC 2005 Conference: Relationship Marketing (Consumer) 74
score represent the share of a person s total category spending that a particular brand could expect to get were there are no market barriers, i.e. if a person could buy what he/she wanted. It deals with commitment: it is a relative share of the space in a person's mind belonging to each brand and is expressed as a value between 0 and 100. The equity score therefore takes into account the level of commitment to other brands in the study, both used (in terms of the total sum of used brands) and non-used (in terms of the relative allocation of the remainder). The overall equity scores for the users in fast food and motor car industries, are 25,8 and 54,7 respectively. Figure 1: Expanded ACSI Model of Fast Food Retailers to Estimate Customer Loyalty as the Ultimate Dependent Variable Table 1: Variance Explained, Reliability and Confidence Values in Fast Food Retailer Data R-square Composite Reliability Lower 95 Upper 95 Lower 99 Upper 99 POQ 0.518 0.905 CE 0.829 PV 0.542 0.930 CS 0.675 0.874 CL 0.450 0.549 0.196 0.692 0.142 0.734 ANZMAC 2005 Conference: Relationship Marketing (Consumer) 75
Figure 2: Expanded ACSI Model of Motor Car Manufacturers to Estimate Customer Loyalty as the Ultimate Dependent Variable Table 2: Variance Explained, Reliability and Confidence Values in Motor Car Manufacturers Data R-square Composite Reliability Lower 95 Upper 95 Lower 99 Upper 99 POQ 0.584 0.909 CE 0.878 PV 0.509 0.939 CS 0.732 0.887 CL 0.368 0.508 0.175 0.602 0.112 0.664 Figure 3: Commitment Categories of the Conversion Model for the Fast Food Industry ANZMAC 2005 Conference: Relationship Marketing (Consumer) 76
Figure 4: Commitment Categories of the Conversion Model for the Motor Car Industry Discussion The findings are of value to both theoretical and managerial perspectives. From an academic viewpoint the major advantage lies in the operationalisation of the two models. The routes followed to reach a perspective on customer retention, is very different in the two models. Customer satisfaction which drives customer loyalty in the ACSI model, does not appear to be a good predictor of customer loyalty. The items employed to measure customer loyalty also scored low on reliability. These findings are in line with the views of various authors who state that customer satisfaction cannot be a sole predictor of customer loyalty. Different items might improve such measurement. This point was raised by a prominent researcher who has applied the ACSI in many studies (Johnson, 2003). Satisfaction is also important in CM. The importance of a brand to a customer as well as the standing of that brand in the consumer s mind, relative to all other competing brands, is a major distinguishing factor of CM. The latter appeal from a theoretical point of view, as this approach is inherent to the conventional models of consumer behaviour which feature the evoked set as central to the consideration of choices. CM categories also offer the advantage that they provide the bases for the formulation of relevant marketing strategies. References American Customer Satisfaction Index, 2001. Methodology report. Ann Arbor, MI : The Regents of the University of Michigan. Anderson, E. W., Fornell, C., Lehmann, D. R., 1994. Customer satisfaction, market share, and profitability: findings from Sweden. Journal of Marketing 58(3), 53-66. Anderson, E.W., Sullivan, M.W., 1993. The antecedents and consequences of customer satisfaction for firms. Marketing Science 12(2), 125-43. ANZMAC 2005 Conference: Relationship Marketing (Consumer) 77
Andreassen, T.W., Lindestad, B., 1998. Customer loyalty and complex services. International Journal of Service Industry Management 9(1), 7-23. Auh, S., Johnson, M.D., 2005. Compatibility effects in evaluations of satisfaction and loyalty. Journal of Economic Psychology 26(1), 35-57. Boulding, W., Kobra, A., Straelin, R. Zeithaml, V., 1993. A dynamic process model of service quality. Journal of Marketing Research XXX (February), 7-27. Clerfeuille, F., Poubanne, Y., 2003. Differences in the contributions of elements of service to satisfaction, commitment and consumers share of purchase: A study from the tetraclass model. Journal of Targeting, Measurement and Analysis for Marketing 12(1), 66-81. Coyles, S., Gokey, T.C., 2005. Customer retention is not enough. Journal of Consumer Marketing 22(2), 101-105. Dick, A. S., Basu, K., 1994. Customer loyalty: Towards an integrated framework. Journal of the Academy of Marketing Science 22(2), 99-113. Edwardson, B., Johnson, M.D., Gustafson, A., Strandvik, T., 2000. The effects of satisfaction and loyalty on profits and growth : products versus services. Total Quality Management 11(7), S917-S927. Fornell, C., 1992. A national customer satisfaction barometer: The Swedish experience. Journal of Marketing 55(1), 1 21. Fornell, C., Johnson, M.D., Anderson, E.W., Cha, J., Bryant, B.E., 1996. The American customer satisfaction index: Nature, purpose, and findings. Journal of Marketing 60(4), 7 18. Gerpott, T.J., Rams, W., Schindler, A., 2001. Customer retention, loyalty, and satisfaction in the German mobile cellular telecommunications market. Telecommunications Policy 25, 249-269. Gustafsson, A., Johnson, M. D., 1997. Bridging the quality-satisfaction gap. Quality Management Journal 4(3), 27-43. Hart, C.W., Johnson, M.D., 1999. Growing the trust relationship. Marketing Management Spring, 8-19. Heskett, J.L., 2002. Beyond customer loyalty. Managing Service Quality. 12(6), 355-357. Heskett, J.L., Sasser, W.E. (Jr), Schlesinger, L.A., 1997. The service profit chain. How leading companies link profit and growth to loyalty, satisfaction and value. New York: The Free Press. Hofmeyr, J., Rice, B., 2000. Commitment-led marketing. John Wiley, Chicester. Iniesta, M.A., Sánchez, M., 2002. Retail-consumer commitment and market segmentation. The International Review of Retail, Distribution and Consumer Research 12(3), 261-279. ANZMAC 2005 Conference: Relationship Marketing (Consumer) 78
Johnson, M.D., 2003. Personal communication. Miranda, M.J., Kónya, L., Havrila, I., 2005. Shoppers satisfaction levels are not the only key to store loyalty. Marketing Intelligence and Planning 23(2), 220-232. Mittal, B., Lassar, W.M., 1998. Why do customers switch? The dynamics of satisfaction versus loyalty. The Journal of Services Marketing 12(3), 177-194. Oliver, R.L., 1999. Whence consumer loyalty? Journal of Marketing 63(4), 33-44. Rice, B., Bennett, R., 1998. The relationship between brand usage and advertising tracking measurements: International findings. Journal of Advertising Research 38 (3), 58-66. Steenkamp, J.-B.E.M., van Trijp, H.C.M., 1996. Quality guidance: A Consumer-based approach to food quality improvement using partial least squares. European Review of Agricultural Economics 23(2), 195-215. Wetzels, M., de Ruyter, K., van Birgelen, M., 1998. Marketing service relationships: the role of commitment. Journal of Business & Industrial Marketing 13(4/5), 406-423. Wold, H., 1982. Systems under direct observation using PLS. In: Fornell, C. (Ed.), A second generation of multivariate analysis: Methods, Praeger, New York, pp. 325-347. Zeithaml V.A., Berry, L.L., Parasuraman, A., 1996. The behavioural consequences of service quality. Journal of Marketing 60 (April), 31-46. ANZMAC 2005 Conference: Relationship Marketing (Consumer) 79