/Statistical-Analysis-of-Marketing-Data

Statistical analysis of a marketing data collected in a super market using python and visualization in order to predict the future outcomes.

Primary LanguageJupyter Notebook

Customer Behavior Analysis Using Descriptive Statistics

Overview

This project focuses on analyzing customer behavior using descriptive statistics and probability distributions. The analysis involves cleaning and preprocessing the dataset, calculating key statistics, exploring probability distributions, and deriving insights to make data-driven recommendations.

Tasks and Deliverables

  • Task 1: Basic CleanUp
  • Task 2: Descriptive Statistics
  • Task 3: Probability Distributions
  • Task 4: Insights and Customer Segmentation
  • Task 5: Conclusion and Recommendations

Major chart outputs

  • Probability Mass Function The chart forecasts the number of purchases each customer is expected to make through various purchasing methods. This allows the store to allocate more resources to the methods with higher customer engagement. Untitled

  • Poisson Distribution These graphs suggest that for all product categories, the most likely scenario is a small percentage change in purchases, either positive or negative. The distributions are symmetric and centered around 0%, indicating that significant changes in purchasing behavior are less probable. Untitled

  • Joint Distribution This graph shows the correalation of money spent by customers with the customer characteristics such as Income, Customer age and Customer relationship tenure. Untitled

  • Average money spent by different category of people This graph shows the probability average money spent on purchases at store categorised by Education, Marital status, kids at home, teens at home. Untitled

Libraries Used

  • pandas
  • matplotlib
  • seaborn
  • numpy
  • scipy.stats

How to Run

  1. Clone the repository.
  2. Run the Jupyter notebook or Python scripts to reproduce the analysis.

Visualizations

  • PMF and CDF plots for purchase categories.
  • Binomial, Poisson, Geometric, and Exponential distribution graphs.
  • Scatter plots and linear regressions for spending analysis.
  • Customer segmentation charts.

Conclusion

This analysis provides a comprehensive view of customer behavior, enabling better decision-making based on data-driven insights.

Author

Ganesh Gouda