Ad Performance Intelligence: Facebook vs Instagram 

              Project Type

Marketing Analytics | Performance Intelligence | Predictive Modeling | Revenue Optimization

Executive Summary

This project analyzed ad performance across Facebook and Instagram to determine which platform delivers better conversion efficiency, cost-effectiveness, and revenue impact.

Using a combination of Power BI, statistical analysis, and regression modeling, the study revealed that Facebook consistently outperforms Instagram in conversion rate, cost efficiency, and revenue generation potential.

A predictive model was developed to estimate conversions based on ad clicks, enabling forward-looking campaign planning rather than purely historical reporting.

The findings support a clear budget optimization strategy: shift investment toward Facebook due to higher ROI and stronger conversion efficiency.

Business Problem 

The business needed to understand:

• Which platform (Facebook or Instagram) delivers better marketing ROI

• How efficiently clicks translate into conversions across platforms

• Whether future conversions can be predicted from engagement data

• How to optimize ad spend allocation across channels

Business Impact (Key Outcomes) 

• Identified Facebook as the higher-performing conversion platform across all months

• Discovered Instagram has consistently higher cost per click and weaker conversion efficiency

• Built a predictive model to estimate expected conversions from click volume

• Provided evidence-based guidance for budget reallocation toward higher ROI channel (Facebook)

• Supported decision-making for conversion-focused marketing optimization

Key Business Insights 

• Facebook consistently delivered higher conversion rates (CVR) than Instagram across the full time period

• Instagram showed higher volatility in engagement and weaker conversion consistency

• Cost per click (CPC) was significantly lower on Facebook, improving overall campaign efficiency

• Engagement metrics (CTR) were similar across platforms, but conversion efficiency differed significantly, showing CTR alone is not a reliable success metric

Statistical & Predictive Analysis 

1. Platform Performance Comparison (Welch T-Test)

A statistical test confirmed that Facebook’s conversion performance is significantly stronger than Instagram’s, validating that the difference is not due to random variation.

2. Conversion Prediction Model (Regression Analysis)

A linear regression model was developed:

Formula:
Conversions = -0.89277 + 0.8182 × Clicks

Interpretation:

• Positive relationship between clicks and conversions

• Higher clicks lead to increased conversions

• Model explains a measurable portion of conversion behavior and can be used for forecasting future campaign outcomes

Revenue & ROI Analysis (Facebook) 

Assuming $5,000 revenue per conversion:

• Total clicks: 67,044

• Conversions: 604

• Revenue: $3,020,000

• Profit: $2,520,000

• ROI: 504% (≈ $5.04 return per $1 spent)

Business Insight:

Facebook ads demonstrate strong scalability potential with high return efficiency under optimized targeting conditions.

Business Recommendations 

• Reallocate advertising budget toward Facebook due to superior conversion efficiency

• Reduce Instagram spend or restructure targeting strategy

• Prioritize conversion-focused optimization over engagement metrics (CTR)

• Run A/B tests to improve ad creative and landing page conversion rates

• Use predictive model outputs to forecast campaign performance before scaling spend

Business Outcome Summary  

This analysis demonstrates that marketing performance should not be evaluated based on engagement alone.

Instead, combining statistical analysis, predictive modeling, and revenue simulation enables:

• Better platform selection

• More efficient ad spend allocation

• Higher ROI-driven marketing decisions

• Data-backed forecasting of campaign outcomes

Tools & Skills

Power BI | Python (Pandas, Statsmodels) | Regression Analysis | Hypothesis Testing | Marketing Analytics | Predictive Modeling | Data Visualization | ROI Analysis

Project Contribution

I independently executed the full analytics workflow:

• Designed data pipeline for ad performance tracking

• Automated data collection using Google Forms and Sheets

• Performed data cleaning and transformation

• Built Power BI dashboards for descriptive insights

• Conducted statistical hypothesis testing (Welch T-test)

• Developed regression model to predict conversions from clicks

• Performed ROI and profitability analysis

• Translated insights into actionable marketing recommendations