e-commerce dashboard (Python& Tableau)

I worked on the Olist dataset - a rich e-commerce dataset -. My mission? To transform this raw, unstructured data into actionable insights using Python for preparation and Tableau for visualization.

Step 1: Data Preparation in Python

- Dataset Collection and Exploration

This initial exploration revealed inconsistencies, such as null values in delivery timestamps and mismatched product categories, which required extensive cleaning.

2. Data Cleaning & Transformation

- Using Python libraries like Pandas, I tackled the following challenges:

- Datetime Standardization: Converted date columns into standard datetime formats.

- Handling Missing Data.

- Data Deduplication: Eliminated duplicate rows in datasets.

- Merging Datasets: Merged Orders and Order Items datasets to create a fact table.

- Data Export: Saved the cleaned and transformed data into Excel files.

Step 2: Data Visualization with Tableau

1. Sales Overview Dashboard:

This dashboard highlights key performance metrics, revenue trends, and geographical insights.

Total Revenue: $16.5M across the dataset timeframe, with notable growth from Q4 2016 to Q4 2018.

Regional Insights: States like São Paulo (SP) and Minas Gerais (MG) were major contributors to sales.

Category Segmentation:

The Pareto Analysis (ABC) showed that ~80% of the profit came from ~20% of the categories, with "Signaling and Security" contributing to a 50% Gross Margin %.

2. Orders Overview Dashboard:

This dashboard focuses on customer behavior, payment preferences, and review patterns.

Order Fulfillment:

Delivery success rate: 97.9%, showcasing a reliable logistics system.

Cancellation rate: 2.1%, indicating minimal friction.

Payment Methods: Credit cards dominated with 77.9K transactions, followed by "boleto" payments at 20K.

Customer Reviews:

A total of 105K items reviewed — with 60K 4-star and above ratings, signaling high satisfaction.