I, chatbot! the impact of anthropomorphism and gaze direction on willingness to disclose personal information and behavioral intentions

Chatbots are becoming increasingly prevalent, serving as customer-service agents on e-commerce sites and personal assistants like Apple's Siri and Amazon's Echo.

This research delves into how two specific features of chatbots—gaze direction and anthropomorphism—affect customer interactions. Although more companies are adopting chatbots, there are mixed results on how best to design and use them. Due to consumer skepticism towards chatbots, it's crucial to understand how to mitigate this skepticism.

Marketing theory often suggests making chatbots as human-like as possible. Based on the humanness-value-loyalty (HVL) model, our study explores how chatbots' nonverbal behaviors influence customers' willingness to share personal information and purchase intentions. We hypothesize that human-like features enhance chatbots' perceived warmth and competence, shaping consumers' expectations of the value of the interaction. As perceived warmth and competence increase, consumer skepticism decreases, leading to greater trust in the service provider. Consequently, customers become more willing to share personal information and show stronger future intentions toward the chatbot.

Through two experiments using real chatbots in simulated shopping environments (car rental and travel insurance), we examine how to make chatbots appear more knowledgeable and empathetic compared to humans.

Our findings indicate that warmth perceptions are influenced by gaze direction, while competence perceptions are influenced by anthropomorphism. These perceptions are key in driving consumer skepticism towards chatbots, which then impacts trust in the service provider, ultimately leading to increased willingness to share personal information and repatronize the e-tailer.

Grounded in the Theory of Mind, our results show that when consumers perceive a chatbot as competent, they are less skeptical, especially if they believe they can understand others' intentions. This research elucidates how to enhance chatbots' human-likeness through appropriate anthropomorphic features and gaze direction.

Our findings emphasize that merely adding digital tools to frontline operations is insufficient to achieve the desired outcomes. Managers need to consider the significant impact of chatbots' characteristics on customer reactions and behaviors.

Specifically, our results can help managers determine the right level of anthropomorphism and the best gaze direction for their chatbots.  Proper adjustments can reduce the risk of poor decisions and increase consumers' willingness to disclose personal information and express positive future intentions towards the chatbot and the company. Moreover, a chatbot’s nonverbal communication through its gaze direction can significantly influence a customer’s likelihood to complete a purchase on the website.

Additionally, our study can help managers reduce consumer skepticism towards chatbots and other digital self-service technologies (SSTs). People tend to apply social rules from human interactions to chatbots based on their anthropomorphism and gaze direction. By making social judgment components like warmth and competence more noticeable, skepticism can be reduced. Lower skepticism enhances the effectiveness of chatbots as marketing tools in service management.

Thus, managers should implement digital assistants with high levels of anthropomorphism, using AI-generated realistic images rather than avatars with lower anthropomorphism levels. Ensuring that these avatars maintain direct gaze with users is also crucial. By fine-tuning these elements, companies can improve customer interactions, fostering trust and encouraging the sharing of personal information, which ultimately benefits the firm's marketing and service management efforts.

Read the full article

Authors at the Department of Management

Gabriele Pizzi – Associate Professor of Marketing

Academic disciplines: Marketing

Teaching areas: Marketing, Retailing

Research fields:  Retailing and Channel Management; Technological Innovation in Retailing

Gabriele Pizzi is an associate professor of marketing and the scientific director of the Observatory on Retailing, which is in partnership with the Retail Institute Italy. In 2017, he received a Research Grant from the Italian Marketing Society for his research on the application of VR in the Retail industry, and he is currently the Principal Investigator for a PRIN project funded by the Italian Ministry for University and Research.