Emotional anthropomorphism of notifications and app engagement: does the usage frequency matter?
Yiling Li, Inseo Hwang, Jeonghye Choi · Journal of Research in Interactive Marketing · 2025
Purpose The purpose of this study is to examine how anthropomorphized notifications with different emotional tones (neutral, positive and negative) influence app engagement in a dog-walking app. Specifically, it explores the differences in user responses between occasional and frequent users. This research aims to fill the gap in understanding how emotional anthropomorphism in app-based marketing communications affects app engagement. Design/methodology/approach Using data from a dog-walking app, this study empirically tests hypotheses regarding the impact of emotionally anthropomorphized notifications on app engagement. Users received notifications with different emotional tones and their engagement, measured by the number of walking days, was analyzed using a random effects negative binomial (RENB) model. Findings Results indicate that anthropomorphized notifications with neutral-toned messages decrease engagement among occasional users, with the decrease in engagement persisting over time. In contrast, frequent users show no significant change in engagement with neutral-toned messages. Positive-toned messages significantly increase engagement among frequent users, while negative-toned messages enhance engagement among occasional users. However, these effects dissipate once the notifications stop, except for the sustained negative impact of neutral messages on occasional users. Originality/value This study extends the anthropomorphism literature by exploring how human-like emotional tones in notifications affect app engagement in a mobile app context. It highlights the importance of user segmentation based on contextual congruence and engagement frequency, challenging the assumption that anthropomorphism universally enhances app engagement and confirming the long-term negative effects of neutral-toned messages on user behavior.