Quality-Variety Tradeoffs in Recommendation Systems on Content Platforms
成果类型：
Article
署名作者：
Zou, Tianxin; Wu, Yue; Sarvary, Miklos
署名单位：
State University System of Florida; University of Florida; Pennsylvania Commonwealth System of Higher Education (PCSHE); University of Pittsburgh; Columbia University
刊物名称：
MANAGEMENT SCIENCE
ISSN/ISSBN：
0025-1909
DOI：
10.1287/mnsc.2025.00300
发表日期：
2026
关键词：
content platforms
content production
hit-driven business
entry
recommendation system
user-generated content
摘要：
User-generated content platforms use recommendation systems (RSs) to deliver content from a vast number of creators to a diverse audience based on content quality and its match to consumers' tastes. Existing research has primarily examined their impact on demand-side factors, such as the quality and variety of content consumption. In contrast, less attention has been paid to their influence on supply-side factors, including creators' entry and quality decisions and their interaction with users' content diets. The literature has also largely overlooked the hit-driven nature of content production-where a few trendy items attract outsized interest in unpredictable ways. We address these omissions by explicitly modeling creators' entry and quality decisions. We reveal a tradeoff between the equilibrium content quality and variety produced on a platform. Although increasing the recommended content's fit to users' preferences benefits consumers and the platform, emphasizing higher quality may reduce content variety and the emergence of trendy items, potentially lowering overall consumer valuation of the platform. We show that whether emphasizing the quality dimension is detrimental depends on consumers' valuations for content match and trendiness, the heterogeneity level of their valuations, creators' capabilities of flexibly choosing their content types, and their barrier to enter the platform. Additionally, consumption variety is a U-shaped function of the emphasis the recommendation system puts on content quality.