OLIST E-COMMERCE DELIVERY ANALYSIS

PROJECT OVERVIEW

This repository contains an exploratory data analysis of delivery performance using the Brazilian Olist e-commerce dataset.

The main focus of the project is to understand delivery time patterns, variability, and potential sources of delay from a customer perspective.

The analysis emphasizes clear reasoning, data validation, and distribution-based interpretation rather than relying solely on single summary metrics.

DATASET

The project uses the Brazilian E-Commerce Public Dataset by Olist, which includes information about orders, delivery timestamps, sellers, and customer locations.

Only delivered orders are used for delivery time analysis. Records with missing or logically inconsistent timestamps are separated and reported instead of being silently removed.

ANALYSIS OBJECTIVES

The analysis is structured into multiple focused objectives:

1. Delivery Time Distribution

   • Understand the typical delivery duration and its variability
   • Define relative fast and slow deliveries using median and interquartile range (IQR)

2. On-Time vs Late Delivery Analysis

   • Compare actual delivery times with estimated delivery dates

3. Review Score between Late Deliveries vs Early Deliveries

   • Explore whether delivery performance between Late Deliveries and Early Deliveries has same average review score or not

Each objective is implemented as a separate notebook to keep the analysis clear and self-contained.

REPOSITORY STRUCTURE

data/
    raw/
        Original datasets
    reports/
        Data quality reports (missing or invalid records)
notebooks/
    00_project_overview.ipynb
    01_delivery_time_analysis.ipynb
    02_on_time_vs_late.ipynb
    03_delivery_review_score_analysis.ipynb

NOTES ON METHODOLOGY

• Data validation is treated as part of the analysis rather than a preprocessing shortcut.
• Missing or logically inconsistent records are separated for transparency.
• Robust statistics such as the median and interquartile range are used to reduce distortion from extreme values.
• Delivery performance is described relative to the observed data distribution rather than external benchmarks.

HOW TO NAVIGATE

If this is your first time viewing the project:

1. Start with notebooks/00_project_overview.ipynb
2. Continue to notebooks/01_delivery_time_analysis.ipynb for the first completed objective
3. Additional notebooks build on this foundation

Kaggle Notebook
Objective 1: https://www.kaggle.com/code/kevinh28/olist-delivery-time-analysis
Objective 2: https://www.kaggle.com/code/kevinh28/02-on-time-vs-late
Objective 3 : https://www.kaggle.com/code/kevinh28/olist-delivery-review-score-analysis