The Complexity of Audiovisual Content in Advertisements Influences User Engagement Behaviours

ABSTRACT Despite the prevalence of audiovisual content on short video platforms, empirical research exploring its impact on user behaviors is still limited. In order to enrich the research related to short video advertisement audiovisual content, we crawl 880 short video advertisements from the TikTok app and creatively apply deep learning and data mining techniques to measure auditory content through sound complexity and visual content through object and texture complexity to explore the relationship between the complexity of audiovisual content and user engagement behaviors. The results show that sound and texture complexity are positively connected with user engagement behaviors, whereas object complexity is adversely correlated. However, this relationship can be influenced by the release date, product type, and saturation of short video advertisements. By firmly linking short video advertisements' audiovisual content to user engagement behaviors, we enhance marketing literature on audiovisual effectiveness. Besides, our feature extraction method significantly enhances the efficiency and accuracy of audiovisual data processing in social media, and the results also aid influencers and businesses in assessing audiovisual content.