THE TRANSITION TO AI CHATBOTS IN RETAIL CUSTOMER SERVICE: AN EXAMINATION OF THE WILLINGNESS TO USE AI CHATBOTS WITH VARYING LEVELS OF ANTHROPOMORPHISM

Abstract

Introduction: The impact that the level of anthropomorphism of an AI chatbot will have on a customer's willingness to use has not been thoroughly explored. Specifically, it is important to understand consumer decision making and willingness to use services in the retail setting. Objective: The primary objective of this study is to reveal how consumers' willingness to use AI chatbots differed depending on their level of anthropomorphism. The study also accounted for independent variables that may influence willingness to use: trust in AI service, trust in the retailer, and perceived usefulness of the AI service. Methods: An online experiment with nine scenarios was used to measure the effect of the level of anthropomorphism of AI chatbots on willingness to use. A convenience sample was used; the study collected 107 viable responses. Pearson Chi-Square, one way ANOVA, and two-way ANOVA tests were used to analyze the data. Results: This study revealed a significant, negative relationship between the AI chatbot's level of anthropomorphism and consumer willingness to use. Trust in AI service, trust in the retailer, and perceived usefulness of the AI chatbot all had significant, positive effects on customer willingness to use the AI chatbot. Discussion: Social exchange theory can be used to explain the results observed in this experiment. The study may provide actional insights for retailers looking to install AI chatbots. Retailers should consider balancing human-like characteristics with practical features when designing their AI chatbots. Additionally, they should increase trust in their technologies as well as their brand through reliable, genuine practices.