Identifying Popular Products at an Early Stage of Sales Season for Apparel Industry

The early phase of launching a new apparel product is critical for gaining insights of its performance and classifying it into different categories such as fast selling, average selling, and slow selling. This information is crucial for optimizing product management strategies and making decisions regarding inventory planning, pricing, and marketing. Many apparel companies rely on rule-based methods conducted by experienced sales managers, which consume significant time and energy from managers and often result in delayed information and low prediction accuracy. We propose a new ranking-based method to identify the product popularity that predicts regional and national rankings of products based on sales data at an early stage of a sales season. Our method enables companies to efficiently identify popular products within a remarkably short span of two to four weeks. To validate its efficacy, we compare the model’s predictions with actual orders from a fashion company in 2021, showcasing a notable 5.9% increase in sales volume when using our approach to guide order decisions.

Funding: J. Wang’s work is supported in part by the National Natural Science Foundation of China [Grants 71931009 and 72171212]. S. Wu’s work is supported in part by the Key R&D Program of Zhejiang Bilateral Industry Joint R&D Plan Project [Grant 2023C04047]. Q. Jin’s work is supported in part by the National Natural Science Foundation of China [Grants 72010107002, 71821002, and 72171212].