Optimizing box content in try-before-you-buy business models with heterogeneous customer groups

Abstract

In online retailing, try-before-you-buy business models are emerging. Customer scan order items to be shipped home to try and decide which items to keep or send back to the retailer. Usually, this business model is combined with personalized recommendations. In this thesis, an optimization model for supporting the decision of what to put in a box to achieve maximum profit in a sales period is introduced. Multiple periods are simulated and overall profit is analyzed. Profit is evaluated by summing the optimum identified in each period. The optimization model treats customer groups differently based on their purchasing behaviours. Policies and business rules are varied to understand the effects on overall profit and customer groups. From the results, managerial implications are drawn to follow a customer-centric approach. To maximize profit, customers with high lifetime value should be treated as preferred if market demand is high and no other factors limit shipping decisions. Notable limitations include inventory availability and market demand. In market situations with limitations, higher-valued customers should be served first. Once a certain scale of the customer base is reached, it should also be focused on other customer groups. It is shown that the prediction accuracy of input data is a crucial concern for sufficiently optimizing box content. Further works could improve the chosen approach to better understand the effect of actions taking place in future sales periods.

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