Textual cues, cognitive load, and social fatigue: Unveiling the reasons behind user discontinuance in conversational AI

Conversational AI is transforming organizational practices across various sectors, significantly reshaping marketing strategies and consumer behaviors. To ensure the continuous use and sustainable development of conversational AI, it is essential to understand the reasons behind users’ discontinuation. This study employs a sequential two-stage mixed-methods approach to explore the drivers of users’ discontinuance in relation to conversational AI. A 2 (dialog Length: long vs. short) x 2 (information volume: high vs. low) between-subjects experimental design is utilized to identify the actual role of the textual cues of conversational AI on users’ discontinuance intentions. Subsequently, a qualitative study is conducted to examine the obstacles and challenges that affect users’ ongoing engagement with conversational AI from the perspective of the customers themselves. The quantitative findings reveal that long dialog and high information in conversational AI, fully mediated by users’ cognitive load and social fatigue, lead to higher discontinuance intentions. The qualitative results further illuminate a set of obstacles and inter-relationships between obstacles that users undergo when discontinuing the use of conversational AI. These findings offer new insights into consumers’ discontinuous usage of conversational AI while also providing practical implications for e-commerce companies and technology providers to implement and improve conversational AI strategies.