Conclusion and Implications

---

In this study, we develop, operationalize and empirically test a model that explains online consumer retention as measured by repurchase. The use of a longitudinal approach provided a better understanding of the causal relationship between satisfaction and repurchase. While previous research on online shopping behavior mainly focused on adoption issues, this study goes beyond adoption to explain the continuation of the behavior; i.e., repurchase. Most earlier studies in the marketing literature on repurchase modeled repurchase as a direct effect of satisfaction. Our findings demonstrated that this was not applicable for online purchase and that a contingency theory was needed; i.e., moderating factors of the link between satisfaction and repurchase should be included. The results of the study presented significant theoretical contribution by showing the importance of considering online shopping habit as moderator of the relationship between satisfaction and repurchase. That is, the link between satisfaction and repurchase can be weakened or strengthened depending on the extent to which a consumer has acquired the habit of online shopping.

Our findings also offered practical contributions. We demonstrated that it is important for cybermarketers to focus their attention and efforts on online shoppers who have already acquired the online shopping habit, as these customers are the ones who are most likely to come back and make repurchases. Furthermore, coupling belief elicitation with prior research allowed us to obtain a salient set of formative measures that resulted in interesting practical implications for managers and cybermarketers about the critical drivers of online customer retention. We identified 13 factors as the main determinants of customer satisfaction, and hence, repurchase. We classified these factors into the dimensions of product satisfaction, process satisfaction and after-sale service satisfaction.

To achieve product satisfaction, "prices of products" was found to be the most significant driver, indicating the importance of pricing strategies in online customer retention. Discounts, coupons and other incentives can be offered to online consumers by making use of the savings in the operational costs resulting from electronic commerce (Gehrke & Turban, 1999).

"Quality of products" came second as the critical factor in determining product satisfaction. As consumers also attach high importance to product quality in evaluating their satisfaction with products, it will be recommendable for managers to seek continuous improvement in the attributes and performance of the products sold on the Internet.

The third significant factor for product satisfaction was found to be "product choices". Cybermarketers should therefore ensure that a wide variety of selections are made available in their Internet storefronts to enhance customer satisfaction. Moreover, consistent with the previous studies, "product description" was found to be another significant driver of product satisfaction. Lohse and Spiller (1998b) found that better product descriptions improved sales significantly. Such impact will even be stronger if product images are added to supplement the descriptions. Similarly, Ho and Wu (1999) found a positive relationship between valuable/accurate product description and online customer satisfaction. Web designers and cybermarketers can therefore enhance the clarity of product descriptions by using pictures when appropriate.

For process satisfaction, several factors were also identified as key drivers. Among these factors, "transaction efficiency" was found to be the most important. Its role is particularly important in improving customer satisfaction with online shopping processes. Similar results were also reported by Ho and Wu (1999), who found that consumers were more satisfied when the quality of the logical support during online transactions was higher. Therefore web designers should make the transactions as efficient as possible by, for example, enhancing the speed of information retrieval and ease of completing online payments.

"Comparative shopping" emerged as the second significant driver of process satisfaction. Most customers expect to be able to compare the prices and availability of products from diverse Internet stores (Rowley, 1996). Web designers should therefore include features that support information search such as a search engine to enable the online shoppers to perform comparison of company products with those of competing brands easily using multiple attributes. Dholakia and Rego (1998) further added that the quality of the information content of web pages could be enhanced by showing the unique advantages of the company products.

The third important driver of process satisfaction was found to be "site accessibility". There will be more website visitors and higher sales if the storefront is accessible from a greater number of "entrances" (Lohse & Spiller, 1998a). Hence, marketers should promote their websites by including their links in other cybermalls, by negotiating reciprocal links with other commercial websites, and by registering their URLs with popular search engines.

"Web page loading speed" was found to be another important determinant of process satisfaction. According to the findings of the multiple surveys conducted by the Graphics, Visualization & Usability (GVU) Center at Georgia Tech, slow loading speed is consistently shown to be one of the key complaints of Internet shoppers. Web designers should hence maximize page loading speed in order to improve customer satisfaction during the shopping process. This can be accomplished by keeping graphics simple and meaningful, limiting the use of unnecessary animation and multimedia plug-in requirements, using thumbnails, providing a "text-only" option, continuously monitoring the server and the Internet routes, and finally, by allowing text to load first, followed by graphics (Gehrke & Turban, 1999).

Our results also indicated that "navigation efficiency" significantly affects satisfaction with Internet shopping processes. Lohse and Spiller (1998a) found that this factor alone explained as much as 61% of sales and 7% of the traffic. With poor navigation, customers may become frustrated, leading to complaints such as, "this is not for computer illiterate people" and, "I had places I wanted to go but couldn't understand how". Web designers should carefully design the online store layout in order to facilitate navigation. Features such as hyperlinks can enable customers to access the relevant information and drill down to more details, when necessary, easily and efficiently. These links should also be well-labeled, consistent and reliable. A map site and an effective search engine would greatly enhance navigation efficiency if there are many underlying links within a website (Gehrke & Turban, 1999).

The last two significant drivers of process satisfaction were "security measures" and "user-friendliness". These two factors had equal importance as indicated by the weights of their formative measures. Security has been acknowledged as a critical factor that affects online shopping behavior in many studies (Gehrke & Turban, 1999; Ho & Wu, 1999). It is therefore important to provide assurance to online shoppers that the transactions are protected. Web designers and marketers should implement security measures such as encryptions, Secure Sockets Layer (SSL), and secure payment systems. Since Internet security is also a matter of perception, security policies should also be clearly communicated to prospective online shoppers by means like adding the statement "Secure Server" to enhance their confidence (Gehrke & Turban, 1999). User-friendliness was also perceived to be of great importance, consistent with all previous studies that stressed the importance of interface design.

For after-sale service satisfaction, we identified two important drivers from the data analysis of the formative measures. "Delivery time" constitutes the most significant factor. Managers of online storefronts should therefore pay attention to the promptness of delivery and ensure offline logistics issues are well handled.

Another driver of customer satisfaction associated with after-sale service was found to be "handling returns". Offering guarantees and a well-defined return policy can effectively enhance the popularity of a website. The implications to web designers are that terms and procedures of returns handling should be clearly stated in the website. Customers should also be provided with feedback channels to make inquiries about the details of these policies.

Approximately 67% of the variance in repurchase remains unexplained. Future research should use more elaborate models incorporating additional antecedent factors beyond satisfaction and online shopping habit. Attempts should also be made in the future to investigate other types of products and industries and incorporate other possible factors such as personal characteristics in models that explain online customer retention.

---