A Comparative Analysis of Customer Sentiment in E-Commerce: Exploring Review Attributes as Indicators of Sentiment Polarity on Amazon and a Manufacturer’s Website

Abstract

E-commerce has been evolving rapidly, greatly influencing consumption patterns and business processes on platforms such as Amazon and manufacturers’ websites. This thesis investigates consumer review sentiment for a specific skincare product across two distinct e-commerce platforms: Amazon and the Manufacturer’s online retail store. Employing both qualitative and quantitative analyses, the study examines review attribute features indicating sentiment polarity in customer reviews. It evaluates indicators of consumer sentiments, including review length, and assesses the influence of factors such as incentivized or Vine-free products, product or brand perception, and overall sentiment polarity. The research utilizes NVivo 14 for sentiment analysis
as well as STATA for logistic regression to identify elements defining consumer sentiment across platforms. Findings reveal noteworthy differences between the two platforms relating to consumer attitudes, with important implications for e-commerce companies aiming to improve customer experience and improve market strategies. By analyzing sentiment polarity in customer reviews, this study postulates guidelines for harnessing customer feedback to boost brand image and increase sales in the competitive e-commerce environment.