Sentiment Signals: Decoding Online Consumer Emotions In India’s E-Commerce Sector

Abstract

The aim of the research is to comprehend and examine the sentiments of the consumers in form of app-based review in the leading Indian e-retailers namely Amazon, Flipkart and Snapdeal, the study. This paper is a study based on the basis of text mining and sentiment analysis that investigates how the customers tend to emotionally respond to their treatment in such sites. The data encapsulation and processing thereof using Python and RStudio and analysis of the data warehouse which has a data set of over 25,000 user reviews collected on Google Play Store were conducted.

It employed a sentiment categorizing method based on lexicon in which a positive, negative sentiment and neutral sentiments along with a more detailed sentiment emotion of trust, joy, anticipation and anger are identified using NRC Emotion Lexicon. Based on thematic analyses of the sentiment trend on each of the three platforms, visualization in the form of word clouds, comparison clouds, and polarity boxplots methodology detailed the content.

This finding reveals that most of the positive messages were received by Flipkart as compared to Amazon and Snapdeal. Ordinary words were used such as service, order and shopping which pointed to the areas of interest to the consumer and words such as good, nice and worst pointed to various emotional reactions. The interpretation reveals that the consumer sentiment is strongly influenced by the how easy the app was to use, the rate at which the delivery was received and fastest, the promotional activity, and the response given by the customer service desk.

Such a study would come in handy in contributing to the generalizing new body of the literature regarding the digital consumer behavior and providing realities of practices at the e-commerce level in the determination of effectiveness on user engagement. It also shows how sentiment analytics is important in competitive benchmarking and brand positioning.