Chatting with a brand spokes-character: A quantitative study on the effects of using brand spokes-characters as virtual chatbot agents

Abstract: This study investigates the consequences of using brand spokes-characters as chatbot agents in customer interactions. Through two studies, the effects of using a brand spokes-character to interact with consumers is compared to using a non-spokes-character chatbot agent. Study 1 investigates the effects during a good chatbot experience where the chatbot is able to understand and give appropriate answers. Study 2 investigates the effects in a poor chatbot experience where the chatbot is unable to give appropriate answers in the interaction. In the good chatbot experience, using a brand spokes-character leads to more favorable brand attitudes compared to using a non-spokes-character through increased anthropomorphizing, conceptual fluency, and brand fit. In the poor chatbot experience in study 2, no significant differences in brand attitude are found due to using a brand spokes-character compared to using a non-spokes-character. Bayesian tests for study 2 indicates that there was no difference in brand attitude between the two groups. Thus, brand spokes-characters are found to offer the opportunity for companies to create more favorable brand attitudes in the context of a good chatbot experience, while not supporting any significant difference in the context of a poor chatbot experience. Despite this, brand spokes-characters are not widely used in the chatbot context. The present study contributes by investigating previously unexplored aspects of chatbot design and finds support that brand spokes-characters should be utilized more in new technological contexts, such as chatbots.