The effect of fixed physical usage patterns on the engagement of physical activity apps: a real-world data analysis

Physical activity applications (PA apps) offer low-cost, time-space-independent interventions that make it possible to promote public health. To increase users' stickiness, the commercially available PA apps usually provide various services to adapt to different app usage patterns of users, thus helping them develop the habit of using apps. However, evidence was rare about whether different usage patterns are associated with the maintenance of PA app use. In this study, we introduce dual process theories and quantify users' app usage patterns in two dimensions: time and space. We analyse the impact of fixed usage patterns on app engagement, by collecting usage data from several commercial-available PA apps in China, which includes 9,175 users. Results show that repeatedly using the PA app at the same time and space could reduce the decline of future app engagement. Moreover, a high degree of self-regulation capacity can mitigate the negative effect of non-fixed usage patterns. This study extends the understanding of health behaviour intervention from the behaviour change level to the behaviour maintenance level. In addition, it provides practical insights for PA app developers in terms of designing behaviour change technologies and for policymakers in terms of facilitating the public's daily physical activities.