A Study on the Service Quality Attributes of AI Banking Chatbots: Based on the KANO Model

With the advancement of Large Language Models (LLMs), chatbots in the financial sector have come to play a variety of key roles, such as handling customer inquiries, providing financial consultations, and supporting transaction processes. This study analyzes the service quality attributes of banking chatbots to improve user experience and service quality, using the KANO model and Timko’s Customer Satisfaction Coefficient. The core elements of AI chatbot service quality were identified, and a survey was designed with seven dimensions and 23 detailed items. Based on the results, each element was systematically classified into attractive, one- dimensional, must-be, and indifferent quality categories. Additionally, the services provided by human bankers were examined to compare and analyze the differences between AI chatbots and human services. Among the service quality dimensions of AI chatbots, elements such as “social presence,” “perceived enjoyment,” and “empathy” are frequently identified as attractive quality (A). However, some elements are also categorized as indifferent quality (I), indicating the need for an in- depth analysis of customer needs for effective improvement. Elements related to “stability,” such as maintaining security and responding to system failures, are deemed the most fundamental aspects of service quality. Furthermore, the ability to accurately understand customer needs, provide professional responses, and ask clarifying questions in cases of uncertainty has become a basic requirement rather than a distinguishing capability. The study also revealed that even within the same quality dimension, elements may possess varying quality attributes.