Foreignness as a double-edged sword for internationalizing cultural goods: deep learning–based semiotic analysis of Hollywood movies in China

Globalization of cultural goods has grown substantively over the past decade. Rooted in the home-country social context, cultural goods are inherently “foreign” when traded in a host country. However, how foreignness is manifested in cultural goods, and how the associated liability and asset of foreignness jointly affect their host-country market performance remain unclear. We address these questions by examining 304 Hollywood movies in China from 2011 to 2018, focusing on semiotic foreignness – the differences in movie synopses and posters between Hollywood and Chinese movies. We utilize Bidirectional Encoder Representations from Transformers (BERT) to analyze synopses and Structural Similarity Index Measure (SSIM) for posters, which allow us to understand how audiences interpret and perceive foreignness in texts and images. The results show an inverted U-shaped relationship between poster foreignness and Hollywood movie box office sales in China, indicating that moderate poster foreignness drives optimal sales. In comparison, box office performance is insensitive to synopsis foreignness. Our study provides valuable insights into the divided views of foreignness and its role in host-country performance. It also addresses the criticism of the poorly specified and operationalized foreignness construct and makes a first attempt to bridge the literature of international business and optimal distinctiveness.