ONLINE SHOPPER'S INTENTION ANALYSIS

INTRODUCTION

Online shopping by consumers is increasing every year, but the conversion rates have stayed relatively stable. For instance, many of us explore e-commerce platforms like Amazon, may add things to our wishlists or shopping carts, but ultimately make no purchases. This reality highlights the necessity for tools and strategies that can tailor promotions and ads to online shoppers and enhance conversion rates. This project will explore multiple factors that influence a buyer's decision.

DATASET

We will be utilizing information from the Online Shoppers Purchasing Intention Dataset for this project, which is accessible through the UCI repository. The primary dataset can be located at this link: https://archive.ics.uci.edu/ml/datasets/Online+Shoppers+Purchasing+Intention+Dataset.

PROJECT MOTIVATION

The motivation behind this project is to address the issue of low conversion rates in e-commerce websites despite the increasing trend of online shopping. The project aims to use data from the Online Shoppers Purchasing Intention Dataset to analyze various factors that influence a purchaser's decision and explore solutions to improve conversion rates. By customizing promotions and advertisements for online shoppers based on their behavior, preferences, and characteristics, the project aims to improve the overall shopping experience and increase sales for e-commerce websites.

PROJECT ACTIVITIES

Exploratory Data Analysis

Univariate Analysis

Baseline Conversion Rate from the Revenue Column 
Visitor-Traffic-Wise Distribution
Analyzing the Distribution of Customers Session on the Website 
Region-Wise Distribution
Analyzing the Browse and OS Distribution of Customers  
Special Day Session Distribution  

Bivariate Analysis

Revenue Versus Visitor Type  
Revenue Versus Traffic Type  
Analyzing the Relationship between Revenue and Other Variables  

Linear Relationships

Bounce Rate versus Exit Rate  
Page Value versus Bounce Rate 
Page Value versus Exit Rate
Impact of Administration Page Views and Administrative Pageview Duration on Revenue
Impact of Information Page Views and Information Pageview Duration on Revenue  

Clustering (Unsupervised Machine Learning algorithm)

Method to Find the Optimum Number of Clusters
Performing K-means Clustering for Informational Duration versus Bounce Rate
Performing K-means Clustering for Informational
Duration versus Exit Rate
Performing K-means Clustering for Administrative Duration versus Bounce Rate and Administrative Duration versus Exit Rate.

CONCLUSION ( summary of our findings and processes/activities )

This embodies the processes or activities and insights acquired during our analysis.
The insights and recommendations are used to make informed business decsions to incrrease output and what have you.

Reference Material

The Data Analysis Workshop By Gururajan Govindan , Shubhangi Hora , Konstantin Palagachev