When Chatbots Struggle with Customer Queries: Can Humans Help Them?

ABSTRACT

Chatbots are becoming increasingly popular for customer service due to their speed and cost-effectiveness. However, chatbots have many limitations; for example, they cannot resolve all consumer queries and rely on human help. This study focuses on two key research questions. First, for what type of queries would customers prefer chatbots to traditional channels like phone calls? Second, when the chatbot cannot resolve customer queries, what is the best way to provide human help so that consumers adopt them in the future? We have used task technology fit and decision point theory to address these research questions. The results of two experiments found that consumers preferred chatbots more for non-complex and non-urgent queries. The results also found that a reduction in decision points led to higher adoption of chatbots in the future. In addition to contributing to theory, this study has significant managerial implications that can help companies effectively employ chatbots.

KEYWORDS:

• Chatbots
• customer service
• task complexity
• time pressure
• task technology fit
• decision points

Acknowledgments

The authors express their gratitude to Mr Parth Hemant Purandare and Mr Davinder Pal Singh Arora, MBA students from SBM, NMIMS University, for their help during the initial stage of this research.

Disclosure statement

No potential conflict of interest was reported by the author(s).