Product Marketing on E-Commerce Websites

Abstract

Online reviews have become the main source of consumer feedback until an informed buying decision is made. Early product reviews seem to have a significant impact on the actual sales of products. By posting reviews on two real-world e-commerce sites in this article, i.e. Yelp and Amazon. We are taking the initiative to examine behavioral characteristics of early reviewers. A person who has posted a review early on considers it as an early reader. We quantitatively classify early users based on their rating trends, the helpfulness ratings earned from others and the association between their feedback and product performance. A person who posted a review early is known to be an early observer. The helpfulness ratings obtained from others and the correlation between their reviews and product results are quantitatively categorized by early users based on their rating patterns. We are presenting a new embedding model based on margins for early divination reviewers by treating the post-review process as a competitive multiplayer game. Extensive studies have shown that our approach outperforms a variety of extreme baselines on two separate e-commerce datasets.