Customer Shopping Behavior Analysis - Portfolio Project

1. Project Overview

This project analyses customer shopping behavior using transactional data from 3,900 purchases across multiple product categories.
The goal is to identify trends in customer demographics, spending patterns, discount usage, product preferences, and subscription behavior to support data-driven business decisions.

This end-to-end project covers:
Python → EDA → Data Cleaning → SQL (MYSQL) → KPI Analysis → Power BI Dashboard → Business Recommendations

---

2. Dataset Summary

• Rows: 3,900
• Columns: 18

Key Features Include:

• Customer demographics: gender, age, age group, location, subscription status

• Purchase details: item purchased, category, purchase amount, season, size, color

• Shopping behavior: discount applied, promo code used, previous purchases, purchase frequency

• Product feedback: review rating

• Logistics: shipping type

• Missing Data: 37 missing values found in review_rating

---

3. Exploratory Data Analysis (Python)

Initial data preparation and cleaning were executed in Python using pandas.

Key Steps

✔ Data Loading & Initial Exploration

• Loaded CSV into a pandas DataFrame
• Used df.info() and df.describe() to understand dataset structure and distributions

✔ Handling Missing Data

• Checked for null values in all columns
• Imputed missing review_rating values using median rating by product category

✔ Column Standardisation

• Renamed all columns to snake_case for readability and consistency

✔ Feature Engineering

• Created age_group by binning customer ages
• Generated purchase_frequency_days using purchase timestamps
• Verified redundancy between discount_applied and promo_code_used
• Dropped promo_code_used due to overlap

✔ Database Integration

• Connected Python to MySQL
• Loaded the cleaned DataFrame into MySQL for structured analytical querying

---

4. Data Analysis using SQL (MySQL)

A series of structured SQL queries were run to answer core business questions, including:

1. Revenue by Gender – Total revenue comparison between male and female customers
2. High-Spending Discount Users – Identified customers who applied discounts yet exceeded average spend
3. Top 5 Products by Review Rating
4. Shipping Comparison – Average purchase amounts for Standard vs. Express shipping
5. Subscribers vs. Non-Subscribers – Revenue and average spend by subscription status
6. Discount-Dependent Products – Top 5 items with highest discount usage rate
7. Customer Segmentation – Classified customers into New, Returning, and Loyal
8. Top 3 Products per Category – Identified using window functions
9. Repeat Buyers & Subscription Likelihood – Subscription patterns among customers with >5 purchases
10. Revenue by Age Group – Identified high-value customer segments

These MySQL queries validated insights discovered during Python EDA and added more focused business intelligence.

---

5. Dashboard in Power BI

The interactive Power BI dashboard includes:

• Key KPIs (Total Revenue, AOV, Discount Usage %)
• Revenue by age group, gender, and product category
• Customer segmentation analysis
• Product performance and review ratings
• Shipping trends
• Discount usage behaviour

The dashboard is designed to be simple, intuitive, and stakeholder-friendly.

---

Dashboard Preview

6. Key Insights

• Customers aged 25–34 generate the highest revenue
• Discount usage has a strong positive effect on purchase amount
• Express shipping customers tend to spend more
• Several high-selling products have below-average review ratings
• Loyal customers contribute significantly to lifetime revenue
• Subscribers show higher average order value than non-subscribers

---

7. Business Recommendations

✔ Boost Subscriptions

Introduce exclusive perks to encourage customers to subscribe.

✔ Customer Loyalty Program

Reward returning customers to increase the number of Loyal customers.

✔ Refine Discount Strategy

Use discounts strategically to boost conversions without hurting margins.

✔ Improve Low-Rated High-Demand Products

Investigate issues with heavily purchased but poorly rated items.

✔ Targeted Marketing Campaigns

Focus efforts on:

• Age group 25–34
• Express shipping users
• Loyal customers

---

8. Business Impact

If implemented, these recommendations could lead to:

• Increased customer retention
• Higher subscription rates
• More effective marketing strategies
• Improved product quality
• Stronger long-term revenue growth

---

9. What I Learned

Through this project, I gained hands-on experience in:

• Python data cleaning & EDA
• Feature engineering
• Writing analytical SQL queries in MySQL
• Customer segmentation and behavioural analysis
• Power BI dashboard development
• Translating insights into actionable business recommendations
• Presenting data in a structured, stakeholder-friendly format

---

10. Future Improvements

• Add machine learning segmentation or churn prediction
• Connect Power BI to a live SQL database
• Perform sentiment analysis on review comments
• Deploy a dashboard web app using Streamlit
• Automate ETL workflows using Python

---

11. Contact Details

If you'd like to connect or have questions about this project:

• Email: onyedikakenechukwu7@gmail.com
• LinkedIn: www.linkedin.com/in/onyedika-kenechukwu-3b3151309

Feel free to reach out always happy to discuss data, analytics, and dashboards!

---