A Comprehensive Analysis of Online Product Reviews

Abstract

Purpose: The rise of e-commerce to prominence can be attributed to the advancement of technology and the internet, which has revolutionised the way businesses are conducted around the globe. In an era of information overload and limitless product options, online product reviews (OPRs) have become a vital source of information. In online shopping, prospective online buyers typically do not have the product experience available to reviewers. Online prospective consumers make their decisions based on the opinions and experiences shared by the buyers cum reviewers in the reviews. The current study aims to investigate the determinants of online product reviews (OPRs) and examine the intricate interplay of online product review determinants in the Indian e-commerce context, which helps consumers navigate through the overwhelming choices available to them.

Methodology: As the market leader in Indian e-commerce, Amazon provides a reliable, fair, and transparent review system. Its standardised, multi-dimensional review format, which covers text, ratings, reviewer attributes, and helpful votes, applies across all categories, making it the most appropriate choice for this study. A multistage sampling technique and exclusion of outliers yielded a comprehensive dataset of 4,900 reviews from 49 best-selling products across three top-selling product categories: beauty, fashion and electronics on Amazon. in. The dataset was analysed using Exploratory Factor Analysis (EFA) in SPSS 22. The extant literature has reported different dimensions of OPRs and analysed single or a limited number of review determinants. Based on prior literature, this study examines seven determinants of OPRs: volume, valence, visual cues, helpful vote count, reviewer expertise, reviewer trustworthiness, and reviewer identity disclosure.

Findings: Exploratory factor analysis identified three latent determinants of OPR: Credibility (reviewer expertise, reviewer trustworthiness, and the presence of visual cues in the reviews), Salience (volume, valence, and reviewer identity disclosure), and Usefulness (helpful vote count). Notably, the Credibility reflects the perceived believability and trustworthiness of the review source and content presentation, which is associated with a higher perceived credibility of the review.  An interesting finding was the negative loading of visual cues, suggesting an inverse relationship with other determinants, indicating that highly credible reviews tend to rely less on images or videos to establish credibility. The salience captures the visibility, prominence and relatability of the reviews. The usefulness reflects peer endorsement and social validation, which enhances the functional value of the reviews. Interestingly, Usefulness appears to be an outcome of other review determinants, with helpful votes aggregating peer judgments of content and source-based reviews. Collectively, these findings provide robust empirical evidence for the complex nature of how consumers evaluate and process online information.

Implications: These findings significantly contribute to the theoretical understanding of online consumer behaviour and electronic word-of-mouth (eWOM) communication. This study empirically validates the multidimensional nature of online product reviews in the Indian e-commerce context. It clarifies how credibility, salience, and peer endorsement or helpfulness interact to shape consumer judgment. Notably, the negative loading of visual cues suggests that images/videos are negatively associated with perceived credibility, aligning with recent studies reporting similar effects. Platforms can better support consumer decision-making by prioritising cues of reviewer expertise and trustworthiness, optimising review prominence and elevating peer-endorsed reviews. Theoretically, the findings extend source credibility, information processing and social proof theories by showing that helpful vote counts aggregate peer judgements across content and source-based cues. Future research should evaluate effects on key outcomes: purchase intentions, sales, and brand image, examine moderating roles of other marketing variables and product categories, and probe the counterintuitive influence of visual cues.