Traditionally, the value of a product has been assessed according to the direct revenues the product creates. However, products do not exist in isolation but rather influence one another's sales. Such influence is especially evident in eCommerce environments, where products are often presented as a collection of webpages linked by recommendation hyperlinks, creating a large-scale product network. Here we present the first attempt to use a systematic approach to estimate products' true value to a firm in such a product network. Our approach, which is in the spirit of the PageRank algorithm, uses easily available data from large-scale electronic commerce sites and separates a product’s value into its own intrinsic value, the value it receives from the network, and the value it contributes to the network. We apply this approach to data collected from Amazon.com and from BarnesAndNoble.com. Focusing on one domain of interest, we find that if products are evaluated according to their direct revenue alone, without taking their network value into account, the true value of the "long tail" of electronic commerce may be underestimated, whereas that of bestsellers might be overestimated.
