The Effects of Stress and Chatbot Services Usage on Customer Intention for Purchase on E-commerce Sites

Abstract

In the rapidly evolving digital marketplace, customer service has become a critical factor influencing consumer behaviour. With the advent of Artificial Intelligence (AI), particularly chatbots, customer service companies are increasingly leveraging technology to enhance user experience. This study explores the relationship between customer emotions, detected during interactions with e-commerce chatbots, and their subsequent purchase intentions. Emotion detection within Human-Computer Interaction (HCI) is a vital area of research, as specific emotions, such as joy or frustration, can significantly impact marketing effectiveness and consumer decision-making. This research aims to understand how emotional responses to chatbot interactions can predict customer's intention to purchase, thereby offering insights for businesses to optimize their AI-driven customer service strategies. The study analyzes four diverse datasets – EmotionLines, CARER, GoEmotion, and EmotionPush – to identify emotion-labelled sentences indicative of purchase intention. Our findings reveal that Neutral and Joyful emotions are predominant in influencing customers' purchase intentions, highlighting the importance of understanding these emotional states in e-commerce settings. While Neutral emotion is most influential, Joy consistently plays a significant role in positive customer engagement. This research underscores the need for e-commerce businesses to focus on emotional intelligence in chatbots, enhancing customer experience and potentially driving sales. Future research directions include examining real chatbot-customer interactions to further understand the impact of AI-driven customer service on consumer emotions and behaviours.