Loading...

Dynamic assortment planning and capacity allocation with logit substitution

Arhami, O ; Sharif University of Technology | 2024

8 Viewed
  1. Type of Document: Article
  2. DOI: 10.1016/j.jretconser.2023.103603
  3. Publisher: 2024
  4. Abstract:
  5. We consider dynamic assortment planning and capacity allocation over multiple periods, where the retailer learns the demand online and actively based on sales data. Our focus is on substitutable products inside a category, where customers' preferences follow multinomial logit, which is unknown to the retailer. This setting, unlike the existing stylish models in retail management, allows us to tackle the real-world problem where the retailer 1) is constrained by a total capacity and 2) replenishes the inventory. Considering a constrained capacity implies that stock-outs may occur, and our algorithm continuously detects and filters out the curtailed demand data to overcome this censorship. Despite the limited capacity, numerical results illustrate that our algorithm has a sublinear regret in various situations. Our analyses also show that the capacity constraint significantly affects the learning and profit of a retailer. Finally, we show that a lower variety in the assortment leads to better revenue when the capacity is small. © 2023 Elsevier Ltd
  6. Keywords:
  7. Demand learning ; Inventory replenishment ; Retail management ; Computer simulation ; Consumption behavior ; Demand analysis ; Numerical model ; Preference behavior ; Profitability ; Retailing
  8. Source: Journal of Retailing and Consumer Services ; Volume 76 , 2024 ; 09696989 (ISSN)
  9. URL: https://www.sciencedirect.com/science/article/abs/pii/S0969698923003545