Review Helpfulness: a comprehensive evaluation of the attributes of user feedback enhancing helpfulness

Abstract

Customer reviews form an important portion of online purchases. With the growing E-commerce industry, customers have turned towards a more informed buying decision process and product reviews have served as a major factor in it. While the effect of reviews on sales of a product has been long proven, we found that there was a lack of comprehensive research on the helpfulness of a review. Our project aims towards providing a comprehensive research analysis on factors which make a review helpful. We performed an analysis of over 2.5 lakh customer reviews classified under two major categories of search and experience goods. We found that the major factors which affected the helpfulness of a review were review length, product type, product ratings, review sentiments and readability of a review. Using our programming model, we were able to find out that the length of a review has a positive impact on its helpfulness. Product type moderates the effect of all the factors i.e., the reviews for search goods were found to be more helpful as compared to experience goods. The helpfulness of a review increased with the ratings given to the product. Reviews with neutral sentiment were found to be less helpful as compared to positive or negative sentiment reviews. Reviews with a moderate readability index were found to be more helpful as compared to reviews which were with too simple or too complex.