Should Chatbot Ads Be Chatty

Powered by artificial intelligence (AI), chatbots are increasingly capable of simulating human-like conversations to enable users to interact with digital devices effectively. But, is this desirable for strategic communications? Will chatbots be more persuasive if they are more human-like, not only in their appearance but also in their interaction and delivery of advertising content? Or will they raise unrealistic user expectations? We explored these questions with a 2 (chatbot profile: human-like vs. machine-like) x 2 (message interactivity: high vs. low) x 2 (ad type: narrative vs. factual) experiment (N = 414). Data reveal that high message interactivity fosters positive attitudes toward the chatbot and the ad by increasing perceptions of social presence, as well as by mitigating violated expectancy. When a chatbot is machine-like in appearance, higher interactivity and adoption of a narrative style of delivery serve to increase ad persuasiveness by heightening social presence. Theoretical as well as practical implications for chatbot advertising are discussed.

Jin Chen 陈瑾
Jin Chen 陈瑾
assistant professor/social science researcher/translator

I study media effects, health communication, emotions, communication technology.