最後修改日期: 2025-05-15
摘要
With the emergence of AI-driven chatbots on digital platforms, maintaining user engagement remains a challenging task for organizations. However, with the rapid rise of chatbots, there still are limitations to understanding the way usability and satisfaction influence long-term user’s retention. This study aims to address this gap by examining the influences of perceived ease of use (PEOU) and user satisfaction on retention intention among the users of two popular chatbot applications, ChatGPT and DeepSeek. A quantitative approach was adopted, and data was collected from 422 respondents via an online questionnaire; 407 valid questionnaires were used for the PLS-SEM Analysis. The constructs were PEOU, Satisfaction and Retention Intention using a 5-point Likert scale. The findings showed PEOU also has significant influence on satisfaction, while the influence of satisfaction on retention intention is very strong. Retention is not directly influenced by PEOU, but paid users have a slightly higher retention. To support these findings, Sentiment Analysis was performed on 10,000 Google Play reviews per app (ChatGPT and DeepSeek) using Phyton and TextBlob. ChatGPT exhibited a greater sentiment score (0.464) compared to DeepSeek (0.329) and had more satisfaction mentions (38.4% vs. 27.7%) and less complaints (3.3% vs. 12.4%). These findings contribute to the role of user satisfaction in driving chatbot loyalty, offering strategic insights for developers and digital strategy teams.
Keywords: Chatbot usability; User satisfaction; Retention intention; ChatGPT; DeepSeek.