A Simulation-based Approach to Dynamic Pricing

By employing dynamic pricing, the act of changing prices over time within a marketplace, sellers have the potential to increase their revenue by selling goods to buyers "at the right time, at the right price." Software agents have been used in electronic commerce systems to assist buyers, but there is limited use of selling agents in today's markets. As dynamic pricing systems become necessary as a competitive maneuver and as market mechanisms become large scale and more complex, there is a growing need for pricing agents to be used to automate dynamic pricing, which challenges sellers to improve their understanding of what are the best agent pricing strategies for their marketplaces. This thesis addresses these issues by presenting the Learning Curve Simulator, a market simulator designed for analyzing agent pricing strategies for a market in which a seller has a finite time horizon to sell its inventory. Through an analysis of several pricing strategies using the simulator, I demonstrate how the Learning Curve Simulator can be used as a tool for understanding the relevant factors in determining an effective dynamic pricing strategy. This simulation-based approach to dynamic pricing demonstrates a technique which can lead to the implementation of dynamic pricing strategies in realworld markets. Thesis Supervisor: Pattie Maes Title: Associate Professor of Media Arts & Sciences This research was funded in part by Mastercard International, the e-markets Special Interest Group, and the Digital Life Consortium.