The Problem
Transforming siloed retail sales data from clients (LGS) into actionable insights for marketing and budgeting strategies.
The Approach
Provisioned a PostgreSQL database via Docker to host raw data. Used Jupyter Notebooks to establish a data connection and employed Pandas for cleaning and exploratory data analysis. Crucially, I integrated an RFM (Recency, Frequency, Monetary) segmentation model to categorize customer behavior.
Technical Stack
Challenges & Constraints
Managing data types across the SQL-to-Python bridge and ensuring the segmentation logic aligned with the client's commercial objectives.
Outcome & Learnings
The RFM model identified high-value clusters and at-risk customers, directly informing a 15% improvement in marketing budget allocation strategies.