Project description

With the rise of artificial intelligence, organizations are increasingly engaging with consumers through automated systems. This project explores how users perceive interactions with text-based dialogue systems, or "chatbots," which use natural language to communicate. Commonly employed in customer service and product advice via websites or messaging platforms, chatbots raise key questions about whether users prefer human agents over automated systems, and how human-like characteristics of chatbots—both verbal and non-verbal—affect user engagement and satisfaction.

To investigate these questions, experimental studies and a meta-analysis were conducted. The experimental findings suggest that while people generally prefer human agents for tasks like study advice, interactions with chatbots were perceived as more enjoyable. Verbal cues that mimic human communication increased the chatbot’s perceived likeability, warmth, and overall satisfaction with the service. However, human-like free-text interactions with chatbots were less appreciated due to perceived issues with user-friendliness.
Chatbots offer significant benefits to both consumers and companies by automating tasks, increasing productivity and boosting customer loyalty. They provide users with round-the-clock access to businesses. However, to be effective and accepted, chatbots need to be carefully developed to ensure they provide real value. The results of this project provide valuable insights to improve the development, design and strategic implementation of chatbots in organizations.

Cooperations

• Prof. Dr. Dominik Papies, Chair of Marketing, University of Tübingen