An Empirical study of Customer Ratings and repeat purchase Behaviour on Blinkit and Flipkart using Regression and Correlation Models

As consumer preferences evolve in India's dynamic e-commerce environment, platforms such as Blinkit and Flipkart must align their operations with customer expectations. This study investigates the impact of customer ratings and reviews on repeat purchase behaviour and monthly sales performance. Unlike traditional studies focusing on delivery speed or pricing, this paper explores how customer-generated ratings influence business outcomes. Regression and correlation models are applied to evaluate these relationships across Blinkit (focused on ultrafast grocery delivery) and Flipkart (a broad-spectrum e-commerce platform). Primary data was collected from 300 regular online shoppers in Mumbai, Bengaluru, and Pune using structured questionnaires and in-app behavioural data. Secondary data was extracted from app reviews, official e-commerce reports, and usage analytics. The independent variables included average product ratings, app interface satisfaction, review length, and review frequency. Dependent variables included repeat purchase frequency and monthly spending.Results from multiple linear regression revealed that for Blinkit, shorter, frequent reviews and higher ratings significantly predicted repeat orders. For Flipkart, the quality of reviews and consistency of ratings showed a stronger correlation with monthly purchase volume. Pearson correlation further confirmed strong positive relationships between rating consistency and consumer loyalty.This study contributes to academic discourse by uncovering rating behaviour as a key influencer in e-commerce growth, especially in post-pandemic consumer patterns. It also provides strategic insights to e-retailers seeking to optimize feedback systems to enhance customer engagement and retention. Further studies could integrate sentiment analysis and machine learning models for predictive purchase behaviour across digital platforms.