INTELLIGENCE IN RETAIL: THE DECISION-MAKING BEHIND CLICK-AND-COLLECT

摘要

Among many web-to-store solutions, we optimize a click-and-collect service, a shopping facility whereby a customer can order products online from a brand's website and collect them from nearby stores. Given a click-and-collect order consisting of a set of products, the problem we present in this paper is to simultaneously select a list of stores from which the customer chooses its preferred pickup point and optimize the profitability. The global profitability comes from several independent factors. A profitable idea is to use the in-store stocks to satisfy all or part of the click-and-collect orders: it avoids transportation costs, it reduces the delivery time, it helps control (speed-up, speed-down) the in-store stocks. A sequence of mathematical programming problems is solved in order to optimize a stochastic criterion that approximates the global profitability. We evaluate our proposed technique with both real and simulated data. The real data come from a large UK retailer now equipped with the click-and-collect tools from devatics.com.