Hi, can you recommend a movie?” Investigating Recommendation Chatbots in Media Streaming Services

Chatbots that interact with users through written natural language have attracted increasing interest in recent years. However, users’ perceptions and adoption decisions of chatbots, especially those for product or content recommendations, have been under-explored. This research focuses on chatbots for content recommendations – in this study called “recommendation chatbots” (RCs) – in media streaming services. Using the example of video streaming services, we examine whether social presence and personalization influence users’ evaluation of a RC in terms of trust, perceived usefulness, and perceived enjoyment. It is further investigated whether a positive evaluation of the RC affects users’ behavioral intentions with regard to RC usage and service loyalty. The preliminary results of the structural equation modelling indicate that perceived social presence only has a positive and significant effect on perceived enjoyment, while perceived personalization was shown to increase users’ trust in the RC, perceived usefulness as well as perceived enjoyment. The results also show that trust, perceived usefulness, and perceived enjoyment positively influence users’ intention to use the RC, which in turn leads to higher service loyalty intentions. Overall, this study enriches the limited research focusing on chatbots for product or content recommendations, providing fruitful insights for research and practice.