Allocation Optimization for Subscription Box Services

The subscription-based models are getting increasingly popular. The most well-known examples are streaming services, food services and monthly subscription boxes. This study considers a hypothetical company with subscription-based customized package as a product. In this type of systems, the demand is for the subscribed packages provided by the company based on the customer preferences, and the company needs to determine the product assortment in the package at each period. We propose two mixed integer programming models for the optimal selection of products to subscription bags while minimizing the total cost of operations or maximizing subscriber satisfaction. We compare cost and satisfaction scores of models to provide insights. We define a general framework for creating a mathematical optimization model for subscription-based services. We segment product by brand value, such as luxury, mid-tier and affordable, to create flexibility and better understand the customer satisfaction and cost. We define a success metric for satisfaction per cost to compare various scenarios with and without product segmentation. We observed that affordable products are a good way to keep the cost down and satisfaction levels adequate.