Although many researchers in information systems and marketing have studied the effect of product reviews on sales, few have looked at their effect on product returns. We hypothesize that, by reducing product uncertainty, product reviews affect the probability of product returns. We elaborate this hypothesis starting with an analytical model that examines how changes in valence and precision of information from product reviews influence the purchase and return probabilities of risk-averse, but rational, consumers. We then empirically test our hypotheses using a transaction-level data set from a multichannel, multibrand North American specialty retailer. Harnessing different consumers’ purchases and returns of the same products, but with varying sets of product reviews over two years, we show that the availability of more reviews and the presence of more “helpful” reviews, as voted by consumers, lead to fewer product returns—after controlling for customer, product, and other context-related factors. Analyzing the purchase behavior of the consumers, we find that when fewer product reviews are available, consumers buy more substitutes in conjunction with a product, potentially to mitigate their uncertainty. Purchase of substitutes, in turn, leads to more product returns. Finally, leveraging a discontinuity in the displayed average ratings, we find that when products are shown with an average rating that is higher than the true rating, they are returned more often. These results support the predictions of our theoretical model—unbiased online reviews indeed help consumers make better purchase decisions, leading to lower product returns; biasing reviews upward results in more returns. The presence of online reviews has important cost implications for the firm beyond the cost of reprocessing the returns; we observe that when consumers return products, they are more likely to write online reviews and that these reviews are more negative than reviews that follow a nonreturned purchase. The online appendix is available at https://doi.org/10.1287/isre.2017.0736 .

Incremental hierarchical text document clustering algorithms are important in organizing documents generated from streaming on-line sources, such as, Newswire and Blogs. However, this is a relatively unexplored area in the text document clustering literature. Popular incremental hierarchical clustering algorithms, namely Cobweb and Classit, have not been applied to text document data. We discuss why, in the current form, these algorithms are not suitable for text clustering and propose an alternative formulation for the same. This includes changes to the underlying distributional assumption of the algorithm in order to conform with the empirical data. Both the original Classit algorithm and our proposed algorithm are evaluated using Reuters newswire articles and Ohsumed dataset, and the gain from using a more appropriate distribution is demonstrated.

C ollaborative filtering algorithms learn from the ratings of a group of users on a set of items to find personalized recommendations for each user. Traditionally they have been designed to work with one-dimensional ratings. With interest growing in recommendations based on multiple aspects of items, we present an algorithm for using multicomponent rating data. The presented mixture model-based algorithm uses the component rating dependency structure discovered by a structure learning algorithm. The structure is supported by the psychometric literature on the halo effect. This algorithm is compared with a set of model-based and instancebased algorithms for single-component ratings and their variations for multicomponent ratings. We evaluate the algorithms using data from Yahoo! Movies. Use of multiple components leads to significant improvements in recommendations. However, we find that the choice of algorithm depends on the sparsity of the training data. It also depends on whether the task of the algorithm is to accurately predict ratings or to retrieve relevant items. In our experiments a model-based multicomponent rating algorithm is able to better retrieve items when training data are sparse. However, if the training data are not sparse, or if we are trying to predict the rating values accurately, then the instance-based multicomponent rating collaborative filtering algorithms perform better. Beyond generating recommendations we show that the proposed model can fill in missing rating components. Theories in psychometric literature and the empirical evidence suggest that rating specific aspects of a subject is difficult. Hence, filling in the missing component values leads to the possibility of a rater support system to facilitate gathering of multicomponent ratings.

We investigate the dynamics of blog reading behavior of employees in an enterprise blogosphere. A dynamic model is developed and calibrated using longitudinal data from a Fortune 1,000 IT services firm. Our modeling framework allows us to segregate the impact of textual characteristics (sentiment and quality) of a post on attracting readers from retaining them. We find that the textual characteristics that appeal to the sentiment of the reader affect both reader attraction and retention. However, textual characteristics that reflect only the quality of the posts affect only reader retention. We identify a variety-seeking behavior of blog readers where they dynamically switch from reading on one set of topics to another. The modeling framework and findings of this study highlight opportunities for the firm to influence blog-reading behavior of its employees to align it with its goals. Overall, this study contributes to improved understanding of reading behavior of individuals in communities formed around user generated content.