Skip to main content

Xingjie Wei & colleagues - Actors' facial similarity and its impact on US movies' box-office performance in East and South-East Asia


Verdiana Giannetti and Jieke Chen from the LUBS marketing department, along with Xingjie Wei have a new paper that sheds light on how the facial similarity of actors in Hollywood movies can significantly affect box office performance in East and South-East Asian (ESEA) countries. The research, published in the International Marketing Review, suggests that casting actors with similar facial features can hinder the audience's ability to recognise and remember characters, ultimately reducing the movie's success in these markets.

The study analysed data from U.S. non-animation movies released between 2012 and 2021, examining their performance in ESEA countries. The researchers employed machine learning face recognition models to analyse the facial image data of the actors, enabling them to quantify the level of facial similarity with a more objective and accurate assessment. The findings indicate that the more facially similar the actors in a movie's cast, the lower the box office returns in the region. This effect is particularly pronounced in countries with lower immigration rates and Internet penetration, as well as in movies with larger casts or less well-known actors.

The researchers propose that Hollywood studios should prioritise casting diverse and easily discernible actors to enhance their movies' performance in ESEA markets. They also highlight the need for greater representation of female actors and those from underrepresented backgrounds, as the study found that facial similarity was significantly higher among female actors across all racial groups.

The study's conclusions emphasise the importance of considering international audiences during the movie production process, particularly given the increasing reliance of Hollywood on global markets. The application of machine learning tools in this research demonstrates the potential for these technologies to provide valuable insights into the factors influencing movie success, paving the way for more data-driven decision-making in the film industry.

The paper can be downloaded here: