Measuring Product Type With Dynamics of Online Product Review Variance

Abstract

The concept of “product type” (experience versus search product) is increasingly important in business research and practice. However, it is not defined or measured precisely in the Internet age due to significantly lower search cost and changes in consumer information search behavior resulting from reliance on information and communications technology.. However, correctly understanding product type has important strategic implications for marketing strategy and digital market design. We take advantage of the greatly available micro level online word-of-mouth data and infer product type (experience versus search product) based on statistical properties of online word of mouth (specifically, online product reviews). We draw on the law of large numbers (L.L.N), and the literature on informational content and online product reviews to analytically propose a mechanism to classify products. Our theoretical analyses indicate that, for a pure search product, when number of reviews (i.e. review sample size) increases as more consumers rate the product, variance of the mean rating will decrease. And for a product with more experience attributes, when number of reviews increases, the variance of the mean rating will not decrease and may instead increase depending on how dominant these experience attributes are. We collect archival data from three different websites (Amazon, Yelp and Ctrip) that collect and publish consumer product reviews to categorize the products and services. Implications of this analytical tool and empirical findings for research, theory and managerial practice are discussed.

Keywords: Online Product Reviews, Product Type, Law of Large Numbers, Information Content, Product Quality, Business Analytics