Digital hoarding behavior of short video platform users: driving mechanisms and complex path analysis based on the C-A-C framework

Purpose This study examines digital hoarding behavior (DHB) among short video platform (SVP) users to reduce unnecessary accumulation, enhance information management efficiency and support SVP sustainability. Design/methodology/approach Guided by the cognition–affect–conation (C-A-C) framework, we analyzed 402 valid responses using partial least squares structural equation modeling (PLS-SEM) and fuzzy-set qualitative comparative analysis (fsQCA). Findings PLS-SEM results show that perceived usefulness (PU), subjective norm (SN) and information overload (IO) significantly increase fear of missing out (FoMO), while PU, time pressure (TP) and SN enhance emotional attachment (EA). Both FoMO and EA promote DHB. fsQCA further reveals three distinct pathways leading to DHB, highlighting multifactorial interactions and configurational patterns. Practical implications This study provides actionable strategies for SVP managers, including optimizing recommendation algorithms, enhancing users’ digital information literacy and implementing intelligent content management tools. Additionally, campaigns and tailored interventions addressing offline and online behavioral differences can promote healthy digital habits, improve information management and reduce unnecessary DHB. Originality/value This study is the first to apply the C-A-C framework to DHB in SVPs, which feature high interactivity and large-scale user engagement. By integrating PLS-SEM and fsQCA, it systematically uncovers the cognitive and emotional mechanisms of DHB, emphasizing its multidimensional nature. The findings contribute to the theoretical understanding of user behavior in high-speed information environments and offer practical insights for fostering SVP sustainability.