Anna-Maria Seeger and Armin Heinzl have presented a paper on the human-like design of chatbots in customer service at the European Conference on Information Systems (ECIS) 2021. The paper with the title “Chatbots often Fail! Can Anthropomorphic Design Mitigate Trust Loss In Conversational Agents for Customer Service?” has won the Claudio Ciborra Award (Runners Up) for the most innovative papers at the conference.
The paper examines the effect of human-like chatbots design on user trust in flawed interactions between customers and chatbots. The results of an experimental study demonstrate that trust loss is significantly lower when customers interact with a human-like chatbot as compared to a computer-like chatbot. In addition the paper reveals the underlying mechanisms that cause this trust shield effect: human-like chatbot signal higher goodwill, which in turn results in higher failure tolerance towards the chatbot. The paper provides important theoretical contributions and practical design guidance.