The Algorithm Vs. The Artist : How do creator-dependent variables affect the view count of animated YouTube videos?

Sammanfattning

YouTube is the world’s second most visited website, and one of the most influential platforms for information spreading and content consumption. As the number of animated YouTube videos continues to grow, partially driven by the increased usage of AI-generation tools, competition for increased algorithmic reach has grown increasingly intense. The purpose of this study was to examine how creator-dependent variables affect the view count of animated videos on YouTube, thus identifying the main drivers for increased algorithmic visibility within the animated video sector. 2,226 animated videos were collected along with 19 creator-dependent variables related to the discoverability, selection and consumption phases of YouTube videos. The variables were examined in relation to view count using negative binomial regression. The results from the regression analysis showed that the majority of variables across all three phases affected view count. However, some variables were found to have no statistically significant effect on view count in both the selection and consumption phase. The findings within the selection and consumption phases were examined in relation to the circumplex model of affect, which suggested that positive valence and high arousal were drivers of increased user engagement. This assumption was supported by some variables, while others suggested that a moderate level of arousal or negative valence was optimal for a higher view count. These findings demonstrate how factors in animated content relate to expected view count. Adjusting variables related to metadata, title, thumbnail and audiovisual content all influence a video’s expected view count. Methodological decisions regarding data collection, data modelling and the omission of potentially relevant variables all serve as limitations for this study.