Assortment optimization under the multi-choice rank list model

: practical application at CurveCatch

Abstract

In today’s highly competitive market, retailers are under significant pressure to determine which products will most effectively satisfy the needs and preferences of their customers to maximize profits given strategical and operational limitations. Most of the assortment planning approaches proposed to help businesses understand customer behaviour are based on discrete choice models. However, many choice models assume that a customer can only purchase at most one product, which in some cases is not an accurate reflection of the real-world purchasing behaviour. In this paper I quantify the benefit of accounting for multi-choice behaviour in rank based choice models and measure the impact that business requirements have on the optimal assortment. Based on the numerical experiment using secondary data provided by CurveCatch, an e-commerce lingerie retailer, I demonstrate that multi-choice modelling significantly improves the revenue generated by the assortment. Furthermore, I provide insight into the implementation of strategic and operational constraints and their impact on the optimal assortment.

Date of Award	2 Feb 2023
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Joren Gijsbrechts (Supervisor)