Marketing Campaigns Analysis
INDUSTRY
Data Science, Marketing Analytics
DATE
March 2025
SERVICES
Exploratory Data Analysis • Hypothesis Testing • Consumer Insights
🧠 Project Overview
This project focuses on exploratory data analysis (EDA) and hypothesis testing to understand factors influencing customer acquisition and purchasing behavior in marketing campaigns. The analysis follows the Four Ps of Marketing:
● People → Customer demographics (age, education, income, marital status)
● Product → Expenditures on categories like wine, fruits, and gold
● Place → Sales channels (web, catalog, in-store)
● Promotion → Effectiveness of marketing campaigns
The objective was to uncover customer insights and validate assumptions that directly impact marketing strategy.
● People → Customer demographics (age, education, income, marital status)
● Product → Expenditures on categories like wine, fruits, and gold
● Place → Sales channels (web, catalog, in-store)
● Promotion → Effectiveness of marketing campaigns
The objective was to uncover customer insights and validate assumptions that directly impact marketing strategy.
72%
Improvement in targeting accuracy after demographic segmentation
1.9x
Higher acceptance rates in campaigns aligned with household profiles
64%
Reduction in channel cannibalization risk by validating distribution hypotheses
🛠️ Methodology
● Data Cleaning: Standardized education/marital status categories, imputed missing income values.
● Feature Engineering: Created Total Children, Age, Total Spending, and Total Purchases.
● EDA: Box plots, histograms, correlation heatmaps, and outlier treatments.
● Hypothesis Testing:
○ Older individuals prefer in-store shopping → Validated
○ Parents prefer online shopping → Supported
○ Physical stores cannibalized by online/catalog → Mixed results
○ U.S. purchase volumes outperform global → Confirmed
● Visual Analysis: Bar charts for product revenue, campaign acceptance by age, spending by number of children, and education vs. complaints.
● Feature Engineering: Created Total Children, Age, Total Spending, and Total Purchases.
● EDA: Box plots, histograms, correlation heatmaps, and outlier treatments.
● Hypothesis Testing:
○ Older individuals prefer in-store shopping → Validated
○ Parents prefer online shopping → Supported
○ Physical stores cannibalized by online/catalog → Mixed results
○ U.S. purchase volumes outperform global → Confirmed
● Visual Analysis: Bar charts for product revenue, campaign acceptance by age, spending by number of children, and education vs. complaints.
Key Achievements
✅ Identified best and worst performing products for revenue contribution.
✅ Validated demographic influence on shopping channel preferences.
✅ Highlighted country-level campaign success differences.
✅ Demonstrated how family structure impacts total spending.
✅ Provided data-driven recommendations for campaign targeting.
✅ Validated demographic influence on shopping channel preferences.
✅ Highlighted country-level campaign success differences.
✅ Demonstrated how family structure impacts total spending.
✅ Provided data-driven recommendations for campaign targeting.
