📊 A/B Testing Analysis: Marketing Campaign Effectiveness

📌 Overview

This project evaluates the effectiveness of an online marketing campaign through A/B testing. We compare two test groups to determine whether advertising influences conversion rates.

• Ad Group: Users who were shown an advertisement.
• PSA Group: Users who were shown a public service announcement.

By analyzing user behavior and applying statistical methods, we aim to determine whether the advertisement significantly increases conversions.

---

🗂 Dataset

• Filename: marketing_AB.csv
• Source: Kaggle
• Key Features:
  • user_id → Unique identifier for each user.
  • test_group → Specifies whether the user was in the Ad or PSA group.
  • converted → Indicates whether the user converted (1) or not (0).
  • total_ads → Number of ads shown to the user.
  • most_ads_day → The day on which the user saw the most ads.
  • most_ads_hour → The hour during which the user saw the most ads.

---

📉 Key Findings

• Conversion Rates:
  • Ad Group: 2.55%
  • PSA Group: 1.79%
• A/B Test Results (Chi-Square Test):
  • Test Statistic: 54.01
  • P-Value: < 0.0001
  • Conclusion: The advertisement significantly increases conversion rates. ✅

---

📊 Visualization

Below is a visualization of the conversion rate comparison:

---

🛠️ How to Run the Analysis

Follow these steps to reproduce the analysis on your local machine.

1️⃣ Clone the Repository

git clone https://github.com/andrewsatyo9/ab_testing_project.git
cd ab_testing_project

🚀 Future Improvements

• Perform multi-variant testing for different ad formats.
• Analyze time-based trends in user behavior.
• Extend analysis with funnel conversion tracking.

---

Author: Andrew Jaya Satyo
LinkedIn: linkedin.com/in/andrew-jaya-satyo-1501992b4
Email: [email protected]