Generative AI and Content Homogenization: The Case of Digital Marketing
(A2026-133815)

AUTHORS

Chaoran Liu, Peking University (UK Campus)

ABSTRACT

Entrepreneurs and small businesses struggle to create high-quality marketing content due to limited resources. Generative AI (Gen AI) offers a powerful solution, but widespread adoption may inadvertently homogenize content across competitors. We examine this tension in the restaurant industry, leveraging Italy’s ChatGPT ban in April 2023 as a natural experiment. Comparing Milan restaurants (treated) to restaurants in four other European cities (control), we find the ban decreased content similarity by 15% (lexical), 12% (syntactic), 2% (semantic), and 3% (language style), demonstrating that ChatGPT access drives substantial homogenization. Critically, reduced homogenization during the ban correlated with 3.5% higher consumer engagement, despite negative effects on posting frequency and content length. Our findings reveal a fundamental trade-off: while Gen AI democratizes content creation, over-reliance risks weakening differentiation and engagement. We provide actionable strategies for businesses to leverage Gen AI efficiency while preserving distinctive marketing voice.