🚀 Ad Campaign Optimization & ROI Measurement

📌 Project Overview

This project analyzes the performance of a digital marketing ad campaign using A/B Testing methodologies. The goal was to quantify the incremental lift in user conversions generated by the ads compared to a baseline Public Service Announcement (PSA) control group.

Using a dataset of 588,000+ user records, this analysis provides data-backed recommendations to optimize marketing spend and scale successful campaigns.

📊 Key Objectives

1. Quantify Impact: Calculate the Conversion Rate and Lift generated by the ads.
2. Validate Results: Use Statistical Hypothesis Testing (Chi-Square) to ensure results are not due to random chance.
3. Optimize Strategy: Segment data by day/time to identify high-performance windows for budget allocation.

##Formula for Lift:

$$\text{Lift} = \frac{\text{Test Rate} - \text{Control Rate}}{\text{Control Rate}}$$

• Statistical Significance: We employ a Chi-Square Test ($\chi^2$) to determine if the observed lift is statistically valid (p-value < 0.05) or due to random chance.

🛠️ Tech Stack

• Language: Python 3.x
• Libraries: Pandas, NumPy, SciPy, Matplotlib, Seaborn
• Techniques: Data Cleaning, GroupBy Aggregations, A/B Testing, Chi-Square Test ($\chi^2$)

📂 Dataset

The dataset contains user interaction data with the following fields:

• user id: Unique identifier for each user.
• test group: "ad" (Treatment) vs. "psa" (Control).
• converted: Boolean (True = Bought, False = Didn't Buy).
• total ads: Number of ads seen by the user.
• most ads day: Day of the week the user saw the most ads.
• most ads hour: Hour of the day the user saw the most ads.

📈 Key Findings & Results

Metric	Control Group (PSA)	Test Group (Ad)
Conversion Rate	1.79%	2.55%
Lift Generated	-	+43.09% 🚀

Statistical Validation

• Test Used: Chi-Square Test for Independence
• P-Value: 1.99e-13 (< 0.05)
• Conclusion: The results are Statistically Significant. We can reject the null hypothesis with >99% confidence.

💡 Strategic Recommendations

Based on the analysis, the following actions are recommended:

1. Scale the Campaign: The ads are driving a massive 43% incremental lift in conversions. The budget should be increased to reach a wider audience.
2. Optimize Day-Parting: Segmentation analysis (see notebook) revealed that [Insert Best Day] has the highest engagement. Future ad spend should be weighted heavily towards these peak days to maximize ROI.

💻 How to Run This Project

1. Clone the repository:

   git clone [https://github.com/yourusername/marketing-lift-analysis.git](https://github.com/ankit2021-nit/Ad-Campaign-Optimization-ROI-Measurement.git)

2. Install dependencies:

   pip install pandas numpy matplotlib seaborn scipy

3. Run the Jupyter Notebook:

   jupyter notebook Campaign_Lift_Analysis.ipynb

---

Created by Ankit Prasad