Impact of helpful reviews on customer purchase intention with special reference to mobile phone reviews

Technological advances in the digital space have provided renewed impetus to businesses. Costly, labor-intensive marketing campaigns have been replaced by digital marketing. However, along with benefits, the increasing sophistication and exponential growth of e-commerce businesses have also introduced new challenges. The large number of similar product offerings and the high volume of reviews have created a technology-induced hurdle for consumers that can impair their thought processes. Often, users will only scan the top few reviews to arrive at a decision. In the current setup, older reviews that accumulate votes over time are found at the top of the helpful review list, in contrast to fresh entrants. The current study proposes placing reviews in appropriate positions in the helpful review list using statistical and scientifically derived helpfulness scores. The study utilized a sample of consumer goods (specifically, mobile phones) and re-ranked reviews based on their expected score. Amazon.in provided the initial review dataset. Random Forest and gradient-boosting regression techniques were used to predict review helpfulness. An Elaboration Likelihood Model was used to explore the impact of central and peripheral cues on review helpfulness. The gradient-boosting regression was the best-performing method of predicting review helpfulness, and the reviews were re-ranked. The re-ranked reviews were tested for helpfulness vis-a-vis the initial ranking of reviews using the survey method. The result indicated that the proposed re-ranking of reviews was more helpful to end users and helped mitigate uncertainty in decisions. The study utilized the Information Acceptance Model to assess the influence of electronic word of mouth on purchase intention.