This summary is provided by the IPR Digital Media Research Center

Dr. Yuan Sun, Dr. Jin Chen, and Dr. S. Shyam Sundar analyzed AI chatbots’ capability to simulate human-like conversations and whether they are more persuasive when they appear human-like in their appearance, interactions, and delivery of advertising content.

An online experiment with 414 adult respondents was conducted from Jan. 10–14, 2021. The participants interacted with either human-like chatbot profiles or machine-like chatbot profiles. Participants were also presented with narrative ads (ads “similar to human ‘storytellers’ which conveyed key messages through stories”) or factual ads (ads which “delivered product details and statistical evidence in a manner that would be consistent with a machine-like presenter”).

Key findings include:

1.) When respondents were able to interact with the chatbot, their attitudes toward the advertisement were more likely to be positive.
— Greater interactivity also led to increased behavioral intention with the advertised product.
2.) Narrative ads enhanced consumer engagement and reduced respondents’ awareness of the advertiser’s intent to persuade.
3.) Narrative ads from interactive, machine-like chatbots generated the most positive attitudes toward advertisements.
4.) Factual ads from less interactive, human-like chatbots generated the least positive attitudes toward advertisements.

Learn more about different types of chatbots and their influence on consumers

Heidy Modarelli handles Growth & Marketing for IPR. She has previously written for Entrepreneur, TechCrunch, The Next Web, and VentureBeat.
Follow on Twitter