Skip to main content

Network structure in electronic commerce

Final Report Summary - ECOMMERCE NETWORKS (Network structure in electronic commerce)

There are numerous networks in e-commerce, including product networks which related products based on shared purchasing patterns. Those 'copurchase networks' are interesting to information technology (IT) economists, since they are endogenously generated by decentralised decisions made by distributed consumers, and thus contain information about the nature of consumer preferences and e-commerce demand. Technology facilitates gathering and analysing detailed data about the structure of these networks, since it is often publicly available in the form of hyperlinks on the products' web pages on e-commerce sites. The set of products that cognitively bounded consumers actually pay attention to is affected by such hyperlinks, and thus they are likely to affect demand. Moreover, these hyperlinks create an informative 'network' of products, one whose properties my research aims to explicitly associate with economic outcomes. There has been much recent research on the phenomena of complex networks across a diverse set of fields. A similar analysis of economic networks will lead to progress in understanding the nature of consumer preferences and demand patterns. My work is grounded on economic theory, although my econometric models also integrate ideas from social network theories and empirical computer science.

Specifically, I pursue in this work four objectives:

(1) to estimate the influence of the network on the demand for products by adapting and extending peer-effect models;
(2) to study the influence of network structure on the distribution of demand, towards providing a deeper understanding of the economic and technological drivers of the 'long tail' of e-commerce;
(3) to develop new demand prediction models that exploit information contained in network structures created as a consequence of choices made by consumers;
(4) to examine the properties of demand networks created by consumer choice, towards relating the structural properties of these networks with alternative models of preference and demand.