To watch or not to watch. The influence of Netflix’s recommendation algorithm on subscription members’ behavioral intentions and recommendation adoption

In this day and age, algorithms have become part of everyday practices. Especially in the growing streaming media industry, algorithmic recommendation systems play an important role in guiding users by suggesting movies and shows to watch. Netflix, the leader in the streaming services industry, has a developed recommender system which challenges viewing practices and is increasingly playing a role in subscription members’ watching behavior. This study aims to investigate how people perceive algorithmic recommenders and to what extent these perceptions are associated with intentions to use the recommender system to find something to watch, as well as users’ actual adoption of the recommendations presented to them by Netflix’s recommender system. Importantly, attitude was examined as a potential mediator in the relationships between perceptions – perceived source credibility and perceived personalization – and intentions to use the recommender system. Recommendation adoption was contextualized by measuring whether people show a higher preference for peer recommendations or recommendations by Netflix. A quantitative survey was distributed among 289 current Netflix users who were 18 years or older. Respondents were mainly found through Instagram, SurveySwap, and Reddit. Mediation analyses using Hayes’ (2017) PROCESS macro in IBM SPSS Statistics 26 showed that perceived source credibility and behavioral intentions are significantly and positively associated, with attitude mediating this effect. Similarly, attitude mediated the relationship between perceived personalization and behavioral intentions. Next, a significant and positive relationship was found between behavioral intentions and recommendation adoption. A paired-samples t-test indicated algorithm aversion, as people showed higher levels of recommendation adoption for peers than for Netflix. The results show that personalization and credibility are key drivers for attitudes, and in turn, attitudes affect intentions and behavior. This study contributes to and expands the current and larger understanding of user perceptions and responses, and the further implications and impacts of this artificial intelligence on society.