This project focuses on A/B Testing and Predictive Analytics to evaluate the success of three different marketing promotions. Using statistical inference and regression modeling, we aim to identify which promotion maximizes sales and understand the impact of variables like Market Size and Store Age.
- Exploratory Data Analysis (EDA): Visualizing sales distributions across different promotions and markets.
- Inferential Statistics: Conducting Two-Way ANOVA to examine the interaction between promotion types and market segments.
- Post-hoc Testing: Performing Tukey’s HSD to pinpoint statistically significant differences between groups.
- Predictive Modeling: Building an OLS Regression model to quantify the influence of each feature on total sales.
- Promotion 1 is the top performer, statistically outperforming Promotion 2 by a significant margin.
- Market Size is a primary driver of revenue; Large markets show a disproportionate increase in sales.
- Store Age has a negligible impact on promotion success, indicating broad campaign applicability.
- Scale Promotion 1 across all Large and Medium markets.
- Discontinue Promotion 2 due to consistent underperformance across all segments.
- Maintain 4-week campaign duration as no significant "Ad Fatigue" was observed.
- Languages: Python
- Libraries: Pandas, Seaborn, Scipy, Statsmodels, Plotly.