Adaptive Change Estimation in the Context of Online Market Monitoring

In the Internet-based economy, the (relative) transparency of e-markets and increasing online market dynamics call for more responsive and encompassing approaches towards the monitoring of markets and competitors. Accordingly, this paper proposes to observe continuously a preselected set of e-commerce Web channels, or online portals, to gather a comprehensive as possible picture of market dynamics. In so doing, a historical market data repository is accumulated based on an adaptive scheme of harvesting Web data online in order to provide dynamic information about both market structure and prices. A description of the proposed estimator for online data sampling based on observed (price) change frequencies is given. Numerical simulations highlight the virtues of the proposed adaptive estimator compared to established Web page change frequency estimators, even more so in case of considering constraints on (observation) resources. As an example, the methodology is applied to the online hotel room booking market.