Quality-adjusted price comparison of non-homogeneous products across Internet retailers

This study compares prices offered by multiple Internet retailers. This task is challenging because e-tailers cannot present their entire assortments to each consumer. Therefore, the quality of the product assortments presented by different e-tailers to each consumer is not directly comparable on an item-by-item basis, resulting in non-homogeneous offerings across retailers. We further consider the interaction between retailers (product information presentation format) and consumers (product information search strategies), which makes price comparisons among the retailers even more non-homogeneous. To grapple with this quality-adjusted price comparison problem for non-homogeneous products, we use a stochastic-frontier hedonic-price regression model to find the “lowest” theoretical price for a product given its characteristics. We then assess the price efficiency of the product as the ratio between this lowest price and the offered market price. This framework allows for the comparison of retailers in their ability to offer the “best deals” even when their actual assortments are not directly comparable in quality. Moreover, this framework provides Internet retailers with a relative measure of price efficiency. This helps them understand when and where they offer competitive prices to consumers. We illustrate our approach empirically in a comparison of price efficiency among three major Internet travel agents on a sample of posted itineraries and airfares. Furthermore, we demonstrate that the price efficiency of an Internet travel agent depends on the format of its website and on consumers' search strategies.