Marketing Analysis

Project Overview

This project analyzes retail customer data to uncover key insights that can support better marketing and business decisions. The analysis focuses on identifying which countries generate the most customers, understanding customer demographics such as age groups, and evaluating the relationship between sentiment scores and product ratings to detect potential inconsistencies.

Additionally, the project examines customer interaction flow across different stages of the website to understand how customers navigate the online shopping experience.

The overall goal is to improve the online shopping journey and develop more targeted, data-driven marketing campaigns.

Objectives

Analyze customer purchase frequency across countries to identify high-value markets.
Examine customer distribution by age group to understand demographic patterns.
Investigate the relationship between sentiment scores and average product ratings to detect meaningful trends or inconsistencies.
Map customer flow and behavior across different stages of the website to evaluate engagement and potential drop-off points.
Develop clear, insightful visualizations to effectively communicate findings to stakeholders.

Tools and Skills Used

SQL: Data extraction, cleaning, aggregation, and complex joins to prepare and organize datasets.
Python: Utilized Pandas for data manipulation and performed sentiment analysis to enrich marketing insights.
Excel / Tableau: Created dashboards and visualizations to present KPIs and communicate analytical findings effectively.

Data Description

The dataset includes customer demographic information, website interaction and engagement metrics, customer ratings and written reviews, and product-level details. Prior to analysis, the data underwent preprocessing to address missing values, remove duplicates, and correct inconsistent formats, ensuring accuracy and reliability throughout the project.

Methods and Analysis

Conducted exploratory data analysis (EDA) to identify key trends, patterns, and anomalies in customer and engagement data.
Segmented customers based on purchase frequency across countries and age groups to understand demographic and geographic differences.
Performed correlation analyses to determine the strongest predictors of customer behavior at each stage of the website journey.
Developed interactive visual dashboards to highlight performance trends and support data-driven decision-making.

Key Findings

Customer Purchase Behavior

Customers in Switzerland make the highest number of purchases, averaging 4 purchases per customer.
Most customers belong to the Adult (25–44) and Middle-Aged (45–64) groups, indicating that these are the primary target age segments.

Rating and Sentiment Analysis

Most products have ratings between 3.0 and 4.0 and sentiment scores between 0.15 and 0.25, reflecting generally neutral to slightly positive feedback.
Climbing Rope stands out as the top-performing product, with the highest average rating and most positive sentiment, showing that customers both rate it highly and express strong positive opinions in their reviews.

Customer Actions Across Website Stages

The highest drop-off rate occurs in the checkout stage, suggesting potential friction before completing purchases.
Click actions are most frequent on the home page, followed by the product page, indicating strong initial engagement but lower conversion in later stages.
Drop-offs on product pages are relatively low, but tracking them over time can help identify emerging issues.

Conclusion

This project demonstrates how marketing and customer data can be transformed into actionable insights. By analyzing customer behavior, segmenting users, and examining trends across each stage of the website journey, businesses can refine audience targeting, optimize marketing budgets, and improve overall conversion outcomes.

The insights generated through this analysis provide a strong foundation for enhancing the online shopping experience and supporting data-driven decision-making.