With dynamic pricing, vendors can adapt retail prices on the fly in order to raise their revenue. This strategy is used in a variety of business domains, including tourism, entertainment and the airline industry. Arnoud den Boer, researcher at Centrum Wiskunde & Informatica (CWI) in Amsterdam studied how to identify optimal pricing strategies to realize maximum revenue. On 7 February 2013 he will defend his thesis ‘Dynamic Pricing and Learning’ at VU University Amsterdam.
Vendors often do not know how to exploit the enormous benefits of dynamic pricing for optimizing revenue. In his research, Den Boer shows the benefits of a deliberate deviation of the apparent optimal price at that moment and experiment with retail prices. “Digital sales data contain valuable information about customer behaviour, especially the relation between demand and retail price. By experimenting with prices, vendors can collect a lot of information about this demand curve,” Den Boer says.
Den Boer designed mathematical algorithms that find the most cost effective way to experiment with retail prices. It sacrifices profits on a short term in order to learn more about customer behaviour and make better pricing decisions in the long term. Den Boer’s pricing algorithm shows how often and in which range vendors should experiment with their retail prices, or in other words, what the optimal balance is between learning and profit maximization.
Arnoud den Boer’s research is part of a recent trend to use Big Data, large amounts of digital information, to make optimal decisions. His results give valuable insights for a large category of decision problems in which Big Data plays a role.
Picture: Cover thesis 'Dynamic Pricing and Learning'