Key points are not available for this paper at this time.
Purpose The study aims to examine the influence of Unified Theory of Acceptance and Use of Technology (UTAUT) and anthropomorphic design cues in determining the level of satisfaction among banking chatbot users. It also tests the moderating impact of the localization of content on the relationship. The study also encompasses expectation confirmation, elucidating the significance of perceived trust in maintaining intention. Design/methodology/approach The study conducted a comprehensive online survey, collecting 667 questionnaires from users of conversational chatbots in both public and private sector banks. We analyse the data using Partial Least Squares Structural Equation Modelling and fuzzy-set Qualitative Comparative Analysis. Findings Performance and effort expectancy, perceived interestingness of interaction and perceived empathy were identified as significant indicators, whereas facilitating conditions, social influence and perceived intelligence were not significant in explaining satisfaction. Perceived trust was a significant mediator, while localization was a significant moderator in all the cases except social influence and satisfaction. Practical implications To improve perceived intelligence and empathy, tech developers should focus on improving the chatbot’s ability to maintain contextual understanding within a conversation where it can remember and reference previous interactions. Future studies might explore the development of banking chatbots that incorporate advanced levels of anthropomorphic characteristics, whether visual or intuitive. Originality/value The work is unique in that it integrates UTATUT, anthropomorphism and expectation confirmation model in the context of conversational banking chatbots, which is not achievable in a single theory-based model. The study also underlined the necessity of localizing chatbot content, recommending that banks engage localized native speakers to help with chatbot training and content creation, where specialists can fine-tune the conversational features.
Misra et al. (Mon,) studied this question.