Estimating Demand for Dynamic Pricing in Electronic Markets

Economic theory suggests sellers can increase revenue throughdynamic pricing; selling identical goods or servicesat different prices. However, such discrimination requiresknowledge of the maximum price that each consumer is willingto pay; information that is often unavailable. Fortunately,electronic markets offer a solution; generating vastquantities of transaction data that, if used intelligently, enableconsumer behaviour to be modelled and predicted.Using eBay as an exemplar market, we introduce a model fordynamic pricing that uses a statistical method for derivingthe structure of demand from temporal bidding data. Thiswork is a tentative first step of a wider research programto discover a practical methodology for automatically generatingdynamic pricing models for the provision of cloudcomputing services; a pertinent problem with widespreadcommercial and theoretical interest