5.1.2 The American Customer Satisfaction Index (ACSI)

The ACSI model, developed in 1994 and illustrated in Fig. 15, builds upon the original SCSB model specification, see figure 14. (For details of the ACSI survey and model see Fornell et al.,1996). The model is estimated for each of the approximate 200 firms in the survey based on a random sample of approximately 250 of the firm's customers. A total of 15 survey questions are used to operationalize the 6 constructs in the model. The survey questions are all rated on 1 to 10-point scales with the exception of price tolerance (described below) and complaint behavior (a dichotomous variable indicating whether the customer has complained or not). In every case, the measurement variables are specified as reflective indicators of the latent constructs in the model.

Figure 15. The American Customer Satisfaction Index model.

The main differences between the original SCSB model and the ACSI model are the addition of a perceived quality component, as distinct from perceived value, and the addition of measures for customer expectations. (By deleting the perceived quality construct and its relationships from figure 15, the reader can readily see the original SCSB model specification in figure 14.)

Quality experts (Deming, 1981; Juran and Gryna, 1988) delineate two primary components of the quality experience, the degree to which a product or service provides key customer requirements (customization) and how reliably these requirements are delivered (reliability). Asking customers to rate customization quality, reliability quality, and overall quality allows the ACSI model to delineate a distinct quality construct that is separate from perceived value. In 1996 the ACSI survey and model were expanded to delineate two general types of perceived quality, product (physical good) quality and service quality. This change was made only for manufacturing durables as they contain both a large product and a large service component. The survey questions used in other sectors to measure perceived quality (customization, reliability, and overall quality) are asked separately for both the product and service aspects of the offering.

The perceived value construct is operationalized using the same two survey questions as in the original Swedish model, a rating of the price or prices paid for the quality received and a rating of the quality received for the price or prices paid. The ACSI model predicts that as both perceived value and perceived quality increase, customer satisfaction should increase. Expected customization and expected reliability were also added to the survey to measure customer expectations using three survey measures (overall expectations, expected customization, and expected reliability).

Fornell et al. (1996) argue that the inclusion of both perceived quality and perceived value into the ACSI model provides important diagnostic information. As the impact of value increases relative to quality, price is a more important determinant of satisfaction. As quality is a component of value, the model also links quality directly to value. There are two measures of customer loyalty in the ACSI model. The first is a rating of repurchase likelihood. The second measure is constructed from two survey ratings: the degree to which a firm could raise its price(s) as a percentage before the customer would definitely not choose to buy from that firm again the next time (given the customer has indicated that he or she is likely to repurchase), and the degree to which a firm would have to lower its price(s) as a percentage before the customer would definitely choose again from that firm the next time (given the customer has indicated that he or she is unlikely to repurchase).

Table 5 Latent and manifest variables ACSI

Table 6 Exogenous and endogenous variables ACSI