Ad Performance Analysis: Facebook Ads and AdWords

Exploring the Impact of Online Advertising through EDA, Hypothesis Testing, and Regression Analysis

📌 Project Overview This project dives deep into the performance of online advertising campaigns run on Facebook Ads and Google AdWords. Using techniques such as Exploratory Data Analysis (EDA), Hypothesis Testing, and Regression Analysis, we uncover insights into ad efficiency, conversion patterns, and campaign effectiveness.

📊 Key Features

Exploratory Data Analysis (EDA):

• Visualizations and summary statistics to understand trends and distributions in the data.

Hypothesis Testing:

• Statistical tests to validate assumptions, e.g., comparing conversion rates across platforms.

Regression Analysis:

Building predictive models to identify key drivers of ad conversions.

Insights and Recommendations:

Actionable insights for optimizing ad performance.

📁 Dataset

The data includes metrics from Facebook Ads and Google AdWords, such as:

• Date : The date corresponding to each row of campaign data, ranging from january 1st 2019, to December 31st 2019.
• Ad Views : The numbers of times the ad was viewed.
• Ad Clicks : The numbers of clicks recevied on the ad.
• Ad Conversions : The numbers of conversions resulting from the ad
• Cost per Ad : The cost associated with running the Facebook ad campaign.
• Click Through Rate : The ratio of clicks to views, indicating the effectiveness of the ad in generating clicks.
• Conversions Rate : The ratio of conversions to clicks, reflecting the effectiveness of the ad in driving desired actions.
• Cost per Clicks : The average cost incurred per click on the ad.

🛠️ Technologies Used

Programming Language: Python

Libraries:

• Pandas & NumPy: Data manipulation and analysis
• Matplotlib & Seaborn: Data visualization
• Statsmodels & SciPy: Statistical tests
• Scikit-learn: Regression models

📈 Analysis Highlights

Exploratory Data Analysis:

• Distribution of monthly/weekly ad conversions.
• Comparison of clicks and conversion rate between Facebook and AdWords.

Hypothesis Testing:

Are Facebook Ads more effective than AdWords for conversions? Does ad spend significantly impact conversion rates?

Regression Analysis:

Predicting conversions based on ad clicks.

🖼️ Visualizations

Here are some key visualizations generated during the analysis:

• Bar plot comparing clicks and conversions across Facebook Ads and AdWords platforms, showcasing their respective performance metrics.
• Monthly or weekly conversion trends for Facebook platform.
• Regression plots predicting conversion rates.

📌 Insights

Key Finding: Facebook Ads have a higher Conversion rate compared to AdWords.

Key Recommendation: Allocate more budget to high-performing campaigns, particularly Facebook Ads, and prioritize spending in May, July, August, September, and November, as these months have a lower cost per ad. Additionally, focus on optimizing underperforming campaigns to maximize ROI.