Comparison of the TCSI models for satisfied and dissatisfied customers
Researchers have claimed that satisfaction levels differ according to gender, age, socioeconomic status, and residence (Bryant and Cha 1996). Moreover, the needs, preferences, buying behavior, and price sensitivity of customers vary (Kutner and Cripps 1997). Previous studies have demonstrated that it is crucial to measure the relative impact of each attribute for high and low performance (satisfaction) (Matzler et al. 2003, 2004). To determine the reasons for differences, a satisfaction scale was used to group the sample into satisfied (8–10) and dissatisfied (1–7) customers.
The research model was tested using SmartPLS 3.0 software, which is suited for highly complex predictive models (Wold 1985; Barclay et al. 1995). In particular, it has been successfully applied to customer satisfaction analysis. The PLS method is a useful tool for obtaining indicator weights and predicting latent variables and includes estimating path coefficients and R2 values. The path coefficients indicate the strengths of the relationships between the dependent and independent variables, and the R2 values represent the amount of variance explained by the independent variables. Using Smart PLS, we determined the path coefficients. Figures 2 and 3 show ten path estimates corresponding to the ten research hypothesis of TCSI model for satisfied and dissatisfied customers. Every path coefficient was obtained by bootstrapping the computation of R2 and performing a t test for each hypothesis. Fornell et al. (1996) demonstrated that the ability to explain the influential latent variables in a model is an indicator of model performance, in particular the customer satisfaction and customer loyalty variables. From the results shown, the R2 values for the customer satisfaction were 0.53 vs. 0.50, respectively; and the R2 value for customer loyalty were 0.64 vs. 0.60, respectively. Thus, the TCSI model explained 53 vs. 50 % of the variance in customer satisfaction; 64 vs. 60 % of that in customer loyalty as well.
Path estimate of the TCSI model for satisfied customers. *p < 0.05; **p < 0.01; ***p < 0.001
Path estimate of the TCSI model for dissatisfied customers. *p < 0.05; **p < 0.01; ***p < 0.001
According to the path coefficients shown in Figs. 2 and 3, image positively affected customer expectations (β = 0.58 vs. 0.37), the customer satisfaction (β = 0.16 vs. 0.11), and customer loyalty (β = 0.47 vs. 0.16). Therefore, H1–H3 were accepted. Customer expectations were significantly related to perceived quality (β = 0.94 vs. 0.83). However, customer expectations were not significantly related to perceived value shown as dotted line (β = −0.01 vs. −0.20) or the customer satisfaction, shown as dotted line (β = −0.21 vs. −0.32). Thus, H4 was accepted but H5 and H6 were not accepted. Perceived value positively affected the customer satisfaction (β = 0.27 vs. 0.14), supporting H7. Accordingly, the analysis showed that each of the antecedent constructs had a reasonable power to explain the overall customer satisfaction. Furthermore, perceived quality positively affected the customer satisfaction (β = 0.70 vs. 0.62), as did perceived value (β = 0.83 vs. 0.74). These results confirm H8 and H9. The path coefficient between the customer satisfaction and customer loyalty was positive and significant (β = 0.63 vs. 0.53). This study tested the suitability of two TCSI models by analyzing the tourism factories in Taiwan. The results showed that the TCSI models were all close fit for this type of research. This study provides empirical evidence of the causal relationships among perceived quality, image, perceived value, perceived expectations, customer satisfaction, and customer loyalty.
To observe the effects of antecedent constructs of perceived value (e.g., customer expectation and perceived quality), customer expectations were not significantly related to perceived value for either satisfied or dissatisfied customers. Furthermore, satisfied customers were affected more by perceived quality (β = 0.83 vs. 0.74), as shown in Table 1. Regarding the effect of the antecedents of customer satisfaction (e.g., image, customer expectations, perceived value and perceived quality), the total effects of perceived quality on the customer satisfaction of satisfied and dissatisfied customers were 0.92 and 0.72. The total effects of image on the customer satisfaction of satisfied and dissatisfied customers were 0.45 and 0.19. Thus, the satisfaction level of satisfied customers was affected more by perceived quality. Consequently, regarding customer satisfaction, perceived quality is more important than image for satisfied and dissatisfied customers. Numerous researchers have emphasized the importance of service quality perceptions and their relationship with customer satisfaction by applying the CSI model (e.g., Ryzin et al. 2004; Hsu 2008; Yazdanpanah et al. 2013; Chiu et al. 2011; Temizer and Turkyilmaz 2012; Mutua et al. 2012; Dutta and Singh 2014). This is consistent with the results of previous research ( O’Loughlin and Coenders 2002; Yazdanpanah et al. 2013; Chiu et al. 2011; Chin and Liu 2015; Chin et al. 2016).
Path estimates of the satisfied and dissatisfied customer CSI model
With respect to the effect of the antecedents of customer loyalty (e.g., image and customer satisfaction), the total effects of image on customer loyalty for satisfied and dissatisfied customers were 0.57 and 0.21. In other words, the customer loyalty of satisfied customers was affected more by customer satisfaction. Customer satisfaction was significantly related to the customer loyalty of both satisfied and dissatisfied customers, and satisfied customers were affected more by customer satisfaction (β = 0.63 vs. 0.14). Consequently, regarding customer loyalty, customer satisfaction is more important than image for both satisfied and dissatisfied customers. Numerous studies have shown that customer satisfaction is a crucial factor for ensuring customer loyalty (Barsky 1992; Smith and Bolton 1998; Hallowell 1996; Grønholdt et al. 2000). This study empirically supports the notion that customer satisfaction is positively related to customer loyalty.
The TCSI model has a predictive capability that can help tourism factory managers improve customer satisfaction based on different performance levels. Our model enables managers to determine the specific factors that significantly affect overall customer satisfaction and loyalty within a tourism factory. This study also helps managers to address different customer segments (e.g., satisfied vs. dissatisfied); because the purchase behaviors of customers differ, they must be treated differently. The contribution of this paper is to propose two satisfaction levels of CSI models for analyzing customer satisfaction and loyalty, thereby helping tourism factory managers improve customer satisfaction effectively.
Fornell et al. (1996) demonstrated that the ability to explain influential latent variables in a model, particularly customer satisfaction and customer loyalty variables, is an indicator of model performance. However, the results of this study indicate that customer expectations were not significantly related to perceived value for either satisfied or dissatisfied customers. Moreover, they were affected more by perceived quality of customer satisfaction. Numerous researchers have found that the construct of customer expectations used in the ACSI model does not significantly affect the level of customer satisfaction (Johnson et al. 1996, 2001; Martensen et al. 2000; Anderson and Sullivan 1993).
Through the overall effects, this study derived several theoretical findings. First, the factors with the largest influence on customer satisfaction were perceived quality and perceived expectations, despite the results showing that customer expectations were not significantly related to perceived value or customer satisfaction. Hence, customer expectations indirectly affected customer satisfaction through perceived quality. Accordingly, perceived quality had the greatest influence on customer satisfaction. Likewise, our results also show that satisfied customers were affected more by perceived quality than dissatisfied customers. This study determined that perceived quality, whether directly or indirectly, positively influenced customer satisfaction. This result is consistent with those of Cronin and Taylor (1992), Cronin et al. (2000), Hsu (2008), Ladhari (2009), Terblanche and Boshoff (2010), Deng et al. (2013), and Yazdanpanah et al. (2013).
Second, the factors with the most influence on customer loyalty were image and customer satisfaction. The results of this study demonstrate that the customer loyalty of satisfied customers was affected more by customer satisfaction. Consequently, regarding customer loyalty, customer satisfaction is more important than image for satisfied customers. Lee (2015) found that higher overall satisfaction increased the possibility that visitors will recommend and reattend tourism factory activities. Moreover, numerous studies have shown that customer satisfaction is a crucial factor for ensuring customer loyalty (Barsky 1992; Smith and Bolton 1998; Hallowell 1996; Su 2004; Deng et al. 2013). In initial experiments on ECSI, corporate image was assumed to have direct influences on customer expectation, satisfaction, and loyalty. Subsequent experiments in Denmark proved that image affected only expectation and satisfaction and had no relationship with loyalty (Martensen et al. 2000). In early attempts to build the ECSI model, image was defined as a variable involving not only a company’s overall image but products or brand awareness; thus image is readily connected with customer expectation and perception. Therefore, this study contributes to relevant research by providing empirical support for the notion that customer satisfaction is positively related to customer loyalty.
In addition to theoretical implications, this study has several managerial implications. First, the TCSI model has a satisfactory predictive capability that can help tourism factory managers to examine customer satisfaction more closely and to understand explicit influences on customer satisfaction for different customer segments by assessing the accurate causal relationships involved. In contrast to general customer satisfaction surveys, the TCSI model cannot obtain information on post-purchase customer behavior to improve customer satisfaction and achieve competitive advantage.
Second, this study not only indicated that each of the antecedent constructs had reasonable power to explain customer satisfaction and loyalty but also showed that perceived quality exerts the largest influence on the customer satisfaction of Taiwan’s tourism factory industry. Therefore, continually, Taiwan’s tourism factories must endeavor to enhance their customer satisfaction, ideally by improving service quality. Managers of Taiwan’s tourism factories must ensure that service providers deliver consistently high service quality.
Third, this research determined that the factors having the most influence on customer loyalty were image and customer satisfaction. Therefore, managers of Taiwan’s tourism factories should allow customer expectations to be fulfilled through experiences, thereby raising their overall level of satisfaction. Regarding image, which refers to a brand name and its related associations, when tourists regard a tourism factory as having a positive image, they tend to perceive higher value of its products and services. This leads to a higher level of customer satisfaction and increased chances of customers’ reattending tourism factory activities.
Is Groupon Good for Retailers?
For retailers offering deals through the wildly popular online start-up Groupon, does the one-day publicity compensate for the deep hit to profit margins? A new working paper, "To Groupon or Not to Groupon," sets out to help small businesses decide. Harvard Business School professor Benjamin G. Edelman discusses the paper's findings. Key concepts include: Discount vouchers provide price discrimination, letting merchants attract consumers who would not ordinarily patronize their business without a major price incentive. These vouchers also benefit merchants through advertising, simply by informing consumers of a merchant's existence via e-mail. For some merchants, the benefits of offering discount vouchers are sharply reduced if individual customers buy multiple vouchers. As a marketing tool, discount vouchers are likely to be most effective for businesses that are relatively unknown and have low marginal costs. Closed for comment; 59 Comment(s) posted.