BEHAVIORAL SHOPPING: ANALYSIS OF CONSUMER BEHAVIOR THROUGH AMAZON PRODUCT LISTINGS

Abstract

Data collected from Amazon.com, reviews and product details, is used to explore various questions. Does a higher list price increase sales for a product? Do better reviews directly increase sales? Are reviewers influenced by other reviews? These subjects have been studied extensively, primarily through controlled experiments. For example, experiments have aimed to reveal higher advertised reference prices affect a consumer’s purchase decisions. We aim to create a framework and dataset that can be used to investigate correlations between various product properties and outcomes. Amazon.com was chosen because of the wealth of reviews, with review author tags, and also product metadata such as sales ranking. This data was collected for the present as well as several points in the past and was analyzed for patterns pertaining to the questions at hand