Learning and Counteracting Dynamic Pricing Strategies
Several e-commerce providers, including plane ticketing and hotel booking websites, present significantly varying prices for the same product. These providers follow dynamic pricing strategies that typically that depend various factors, such as a users geographic location (higher prices in richer countries), a user’s IP address (lower prices for comparison websites), or even a user’s cookies (higher prices for users that already checked the price in the past).
The goal of these strategies is to maximize profits by always charging the highest price the customer is willing or able to pay. Since different customers might not all be willing to pay the same price, the website providers aim to determine the optimal price for each customer, based on the available information. Many of these parameters can be influenced by the user, e.g., by using proxies or VPN connections to change the IP address or modifying the header or the browser history. While such pricing strategies result in a situation where rich or loyal customers potentially pay significantly higher prices and customers that intensively look for low prices find a much better price for exactly the same product, these strategies could be exploited to automatically retrieve the lowest available price for a given product.
We propose two master's thesis projects for learning and counteracting dynamic pricing strategies. These two projects would be suited for two students that work in close collaboration and are listed above.
Contact: Esfandiar Mohammadi