Market intelligence and price adaptation

In the context of e-commerce, it is critical for a retailing company to be able to assess the market and respond quickly to changes in competition and/or service levels and availability of its products. If the company operates globally, and the geographical constraints on the supplier are different from those on the consumer, it is even more crucial to assess each market with respect to its characteristics and dynamically price every individual product accordingly. In this paper we report on an agent-based distributed system that was developed to support the retailing operations of one of the largest online book sellers in the UK, which has millions of books on offer and operates internationally in a number of market places. The system supports the collection of market data, processing of this data, and applying dynamically one of set of predefined pricing patterns. We describe the principles on which the price adaptation is based, the (long-running and proven successful) distributed scalable software architecture that supports it, illustrate how the mechanism copes with some typical situations system that arise and highlight some of the lessons learned from the development experience