From text boxes to talking faces: Comparing chatbots and digital humans for online review collection

Digital humans (agents controlled by artificial intelligence with highly realistic faces and voices) are beginning to appear in place of text-based chatbots in routine customer service roles, such as soliciting feedback and reviews. There are two fundamental ways digital humans may transform the process of providing online reviews. First, with highly human-like visual appearances, digital humans can significantly enhance the perceived humanness of agents collecting reviews compared to traditional text-based chatbots. Second, by enabling spoken rather than typed communication, the interaction may take on a more casual, conversational nature, making the process feel like an informal conversation. We conducted an online experiment to investigate how users respond to digital humans and text-based chatbots when providing online reviews of restaurants. Our findings suggest that digital humans enhance the perceived humanness of the agent, making the review process feel more like a casual conversation. This perceived humanness and casual conversational nature mediate the impact of digital humans on outcomes, increasing perceived effectiveness, efficiency, satisfaction, and usage intention. Our study demonstrates that replacing a text-based chatbot with a verbally interactive digital human has medium to large effects on improving key outcome measures, suggesting that deploying digital humans can help firms increase the volume of online product and service reviews.