Understanding Buyers' Behaviors in Vertical E-Commerce with Limited-edition Merchandise.

Personalized recommendation is one of the keys to provide a successful e-commerce user experience. To offer such recommendations, the use of historical data and product metadata is critical. However, recently emerging vertical e-commerce that focuses on a specific product category faces challenges in making personalized user experience due to rapid product life cycle and few historical data. This study aims to understand user behavior in this new challenging vertical e-commerce environment with a real-world vertical e-commerce platform called Twenty. We analyzed Twenty's three-month order data. The results showed that generally buyers are not following popular sellers and have their own preferences.