Data & Analytics
E-Commerce Customer Insights (Olist)
Python, SQL
Queried 100K+ order records to engineer delivery KPIs and identify correlation between shipping delays and low ratings. Text analysis on customer complaints for service patterns.
Overview
Analysis of Olist Brazilian e-commerce dataset (100K+ orders). Built delivery KPIs, explored relationship between shipping delays and customer ratings, and performed text analysis on reviews.
Approach
Used SQL for aggregation and joins, Python (pandas) for statistical analysis and NLP on review text. Created metrics for on-time delivery, average delivery time by region, and sentiment patterns.
Results
Identified strong negative correlation between shipping delays and ratings. Uncovered common complaint themes to prioritize operational improvements.