The Impacts of Information-Seeking Strategies and Social Presence on the Interaction Between Customers and Conversational Agents
Files
Date
Authors
Supervisor
Item type
Degree name
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Many organisations in different industries have jumped on the bandwagon and implemented conversational agents (CAs) to support customer service operations. Although CAs are found to offer many benefits to their customers, they also pose various challenges, which raises several concerns about their effectiveness. Many CA implementation projects have failed to meet initial expectations. This high failure rate in CA implementation indicates a lack of understanding about customer-CA interactions. Therefore, it is necessary to understand how the interactions take place and why many of them fail, in order to close the gap between organisations’ expectations and the actual performance of CAs. This study takes a close look at the role of information-seeking strategies and social presence in shaping the interaction process by investigating the nature of customer-CA interactions and their success and failure factors. The following research question was addressed: “How do information-seeking strategies and social presence shape the outcome of interactions between customers and conversational agents?” Informed by the social information seeking model and the concept of social presence, this research examined the effects of information-seeking strategies and social presence on how interaction outcomes are shaped. The analysis involved investigating the interaction logs of 507 conversations between customers and a chatbot, which is a form of CAs that communicate with customers using text-based messages. The chatbot was implemented on the website of an electric power company to serve customer online services. By taking an abductive qualitative research approach and adopting a configurational thinking methodology, the researcher identified five distinct information-seeking strategies that customers used to interact with CAs: complete-sentence strategy, fragmented strategy, keyword search strategy, FAQ strategy and social reciprocity strategy. These strategies represent configurations of four interrelated factors that influence the strategy selection: context-related factors, goal-related factors, information-related factors, and technology-related factors. The selected strategies were found to reciprocally interact with the degree of social presence to form outcomes. Factors involved in the interplay are indicated as those that contribute to the success and failure of the customer-chatbot interactions. This study proposes a process model that illustrates the interplay between information-seeking strategies and social presence in customer-chatbot interactions. This explains the intricate interplay of the various factors in shaping the interaction outcomes.