Chatbot Interactions: The impact of customer confusion on customer experience

Abstract

This study aims to understand the effects of customer confusion on customer experience (CX) outcomes: satisfaction and loyalty, in a chatbot environment. In order to achieve this objective, the study employs a Wizard of Oz (WoZ) approach where customers interact with a simulated shopping chatbot named Cody in a shopping scenario. The experiment utilizes error messages customized for different conditions to induce confusion while measuring customer sentiments through sentiment analysis to comprehend their emotional responses during this service encounter. The findings of this research indicate that confusion negatively impacts customer satisfaction and loyalty but is entirely mediated by their feelings and behavior. Specifically, hedonic value and seeking additional information play critical roles in mediating the impact of confusion on CX outcomes. Thus, empathetic chatbots that address customers' emotions are imperative for improving CX in businesses. Moreover, online shopping experiences indicate that longer durations can positively influence satisfaction and loyalty; hence businesses must focus not only on reducing confusion but also managing customer feelings for improved CX. It is crucial to acknowledge the limitations of the study's sample representativeness and control over participants' responses; hence future research should replicate the experiment with more diverse samples while exploring the relationship between feelings and behavior concerning CX outcomes. In summary, this research contributes to our comprehension of how confusion affects CX and offers useful suggestions for enhancing customer experiences in chatbot settings.