From Service Failures to Service Recovery: Measuring Online Consumers' Emotions During the Use of Online Product Configurators

Over the past decade, business-to-consumer electronic commerce has seen tremendous growth, opening up vast opportunities for retailers to operate on a global scale. In Hong Kong, small and medium enterprises (SMEs) are increasingly participating in businessto- consumer (B2C) e-commerce to expand their geographical reach, better target customers, or harness synergies between their online and offline businesses. Yet, participating in e-commerce can also have significant drawbacks. One problem is the online retailers’ limited ability to respond to customers’ problems during online shopping tasks (and specifically, during difficult or complex tasks, such as using product configurators). Such service failures have been shown to contribute to negative emotions (e.g., anger and frustration), which can lead to the termination of transactions, negatively influence repurchase intentions, and increase people’s tendency to spread negative word-of-mouth.At the same time, effectively addressing the customers’ problems and recovering from such service failures can have positive effects on customer retention and positive wordof- mouth. In a face-to-face, offline retailing context, detecting customers’ negative emotions is relatively easy, and appropriate action can be taken to remedy the situation and retain the customer. However, online retailers lack the ability to directly observe their customers, and need to find other ways to infer customers’ emotions.The proposed study will attempt to discover ways to detect users’ negative emotions in online environments, so as to mitigate negative effects of e-service failures. To this end, we will bring together research from human-computer interaction, marketing, and psychology, and hypothesize about how users’ emotions can be detected in online contexts. Further, we will develop a system to infer users’ negative emotions “on the fly;” such system can be used to provide users with additional help or guidance if negative emotions are detected. We will test the efficacy of the system in different cultures, so that online businesses targeting global customers can benefit from our findings. The proposed study will have some important long-term implications, as it will help online retailers compete in the global marketplace; in addition, the results will serve as a foundation for further research into detecting negative emotions in other contexts, such as in mobile applications. Overall, given the increasing importance of the Internet for Hong Kong’s economy, this study will be of great importance to businesses in Hong Kong as well as mainland China.