Visual Attributes of Thumbnails in Predicting YouTube Brand Channel Views in the Marketing Digitalization Era

Ha Eun Jang,Seung Ho Kim,Jong Seok Jeon,Joo Hee Oh
DOI: https://doi.org/10.1109/tcss.2023.3289410
2023-01-01
IEEE Transactions on Computational Social Systems
Abstract:Digitalization has shaped the relationships among companies, consumers, and social participants with the advancements of digital technology. As a result, companies are increasingly adopting and integrating the YouTube system for content dissemination and engagement with their customers and stakeholders. This study proposes a prediction model that analyzes 16 278 image datasets collected from 137 brand channels’ videos with over 100 000 views, generated before September 26, 2022. Using the dataset, we analyze the factors affecting the number of content views of brand channels and construct a view prediction model. This study finds that the characteristics of the thumbnail image, offline top brand characteristics, and channel size (number of subscribers and number of channel videos) significantly influence YouTube’s online channel views. The results of this study provide a strategy for brands to communicate more actively with stakeholders through the YouTube system.
What problem does this paper attempt to address?