From visuals to value: leveraging generative AI to explore the economic implications of movie poster

This study introduces a novel exploration into the impact of visual elements on consumer behavior in the film industry, utilizing generative AI. By employing expectancy violations theory, this study examines how the mood and tone suggested by movie posters—a key visual advertising tool—contrast with the actual content of the films, revealing a significant influence on consumer decisions. Through feature extraction from movie posters using generative AI and regression model analysis, this study demonstrates that deviations from audience expectations, as suggested by movie posters, positively affect box office performance. However, this impact varies depending on whether the movie is produced by major or non-major studios, with major studio productions benefiting more from mood and tone congruence. This study extends existing literature by highlighting the role of visual cues in movie posters and offers practical insights for movie marketers using expectancy violations to enhance audience engagement and box office performance.