We study social influence in an online music community. In this community, users can listen to and “favorite” (or like) songs and follow the favoriting behavior of their social network friends—and the community as a whole. From an individual user’s perspective, two types of information on peer consumption are salient for each song: total number of favorites by the community as a whole and favoriting by their social network friends. Correspondingly, we study two types of social influence: popularity influence, driven by the total number of favorites from the community as a whole, and proximity influence, due to the favoriting behavior of immediate social network friends. Our quasi-experimental research design applies a variety of empirical methods to highly granular data from an online music community. Our analysis finds robust evidence of both popularity and proximity influence. Furthermore, popularity influence is more important for narrow-appeal music compared to broad-appeal music. Finally, the two types of influence are substitutes for one another, and proximity influence, when available, dominates the effect of popularity influence. We discuss implications for design and marketing strategies for online communities, such as the one studied in this paper.

Through reimbursing a portion of the transactional amount to some consumers in a form of cash back, merchants are able to exercise third-degree price discrimination by offering two asymmetric prices via an online dual channel. To better understand such a novel pricing mechanism, we develop a game theoretical model and start our analyses with a market consisting of one merchant, one affiliate site, and consumers heterogeneous in their product valuation. From a price point of view, cash-back shopping appears to provide site users with a saving opportunity since the effective post-cash-back price they pay is perceived to be lower than the regular price targeted at nonusers. However, we find that under some conditions, this seemingly lower price could be actually higher, compared with the optimal uniform price when the merchant does not price discriminate. An important implication is that all consumers may end up suffering from higher prices in the presence of the cash-back mechanism. This surprising result, referred to as the cash-back paradox, defies a common intuition that a price-discriminating firm must raise the price for one segment of consumers but decrease it for the other. We also develop two extensions to seek explanations behind various industry practices. We find that it is in a merchant’s best interest to affiliate with multiple sites, and the resulting competition improves overall market efficiency. Moreover, merchants who are disadvantageous in brand valuation should target price-sensitive consumers by strategically offering cash-back deals. Our results, consistent with several real-world observations, have useful implications for marketers. The online appendix is available at https://doi.org/10.1287/isre.2017.0693

Abstract-With the advent of the World Wide Web, many business applications that utilize data mining and text mining techniques to extract useful business information on the Web have evolved from Web searching to Web mining. It is important for students to acquire knowledge and hands-on experience in Web mining during their education in information systems curricula. This paper reports on an experience using open Web Application Programming Interfaces (APIs) that have been made available by major Internet companies (e.g., Google, Amazon, and eBay) in a class project to teach Web mining applications. The instructor's observations of the students' performance and a survey of the students' opinions show that the class project achieved its objectives and students acquired valuable experience in leveraging the APIs to build interesting Web mining applications.

No abstract

No abstract