This project focuses on exploratory data analysis (EDA), advanced analytics, and visualization to understand customer churn patterns in the telecom industry. The analysis is conducted using Python, AWS S3, Snowflake, and Power BI, along with comprehensive business documentation for insights and decision-making. https://drive.google.com/drive/folders/1CqmgRNK2ZsTivxgx3OUi2-Sl2qLfmLIU?usp=drive_link This Link has all the Project related files. Tools Used: Python, AWS S3, Snowflake, Power BI
I led a comprehensive data analysis project to identify key factors influencing customer churn in a telecom dataset, using Python for exploratory data analysis (EDA), AWS S3 and Snowflake for cloud-based data storage and querying, and Power BI for dashboard creation and reporting.
Approach & Methodology: 🔹 Univariate Data Analysis: Used histograms to understand the distribution of key variables such as tenure, charges, and service usage. Created count plots to visualize categorical features like contract type, payment method, and their relation to churn. Applied box plots to observe spread and detect outliers in numerical variables grouped by churn status. 🔹 Bivariate Data Analysis: Employed scatter plots to explore relationships between variables like monthly charges vs. tenure. Used box plots to compare continuous features across churn and non-churn groups. 🔹 Statistical Testing: Performed Chi-Square tests to assess relationships between categorical variables (e.g., contract type vs. churn). Applied ANOVA tests to identify significant differences in numerical variables across multiple customer segments. 🔹** Multivariate Analysis:** Conducted correlation analysis to detect patterns among multiple continuous variables and uncover key drivers of churn. Cloud & Visualization Tools: Stored and queried the dataset using AWS S3 and Snowflake to ensure scalability and real-time access to updated data. Built an interactive Power BI dashboard to display KPIs like churn rate by customer segment, service usage trends, and tenure vs. contract type, helping stakeholders easily track churn risks. Outcome: The analysis provided data-backed recommendations to improve customer satisfaction and retention. These included tailoring contract offerings, targeting high-risk segments with personalized offers, and improving services linked to higher churn. I also created a comprehensive report summarizing findings and presented it using visuals and business-focused insights.