Essays on Novelty in Online Reviews

The popularity of electronic word-of-mouth (e.g., online reviews and ratings) has increased dramatically over the years. Consequently, it has become important to understand the various ways in which online reviews impact different stakeholders, such as consumers, businesses, platforms, and reviewers. Despite the efforts to understand the relationship between online reviews and consumers’ purchase decisions, a surprising omission in the existing research is the impact of new information provided in online reviews. Within this context, my dissertation consists of three essays that focus on the impact and the use of novel information in online reviews. In the first essay, I examine how novel information in reviews influences restaurant check-ins and review helpfulness and how this relationship changes as review volume changes. I show that novel information positively impacts the helpfulness of reviews and restaurant check-ins—however, the impact becomes negative when the review volume becomes high. To provide a better understanding of the relationship between novel information in reviews and restaurant check-ins, I do a mediation analysis and show that review helpfulness partially mediates the effect of novel information on restaurant check-ins. The indirect effect (through helpfulness) is moderated by review volume. I also show that the impact of novel information on check-ins changes with restaurant type, with the impact higher for high-priced restaurants. In the second essay, I focus on a common problem that review platforms face when a large number of reviews are posted for a product or service. Because readers are typically unable or unwilling to read all posted reviews, platforms aim to identify a small subset of reviews that can provide enough information to help readers make a decision. To address this problem, I present a formulation to determine the novelty of a review subset and propose an efficient review selection algorithm to maximize the amount of novel information in a review subset given the subset size. The proposed approach greatly improves the amount of novel information relative to benchmark approaches. I design heuristics for real-time environments and show the proposed approaches can be extended to scenarios preserving the average opinion and aspect coverage in the corpus. The third essay examines the moderating role that credibility can play on the effect of novelty on review helpfulness. This is motivated by prior studies from the psychology literature that have shown the impact of credibility on the perception of information that people receive in traditional environments. Research on review platforms has identified two types of credibility characteristics: source credibility (e.g., a platform-designated status of a reviewer) and rating credibility (closeness of a review’s star rating to the product’s average star rating). I examine the interaction between the two credibility characteristics (reviewer’s source credibility and review’s rating credibility) and a review’s novelty on a review’s helpfulness. I show that while source credibility has a substitution effect on the impact of the novelty of a review on its helpfulness, rating credibility has a complementary effect on the impact of novelty.