Data Storytelling Dashboard, E-Commerce Analytics

An interactive data visualization and analytics dashboard that transforms raw e-commerce data into actionable business insights.
Built with Python, Streamlit, and Plotly, this project demonstrates advanced data storytelling, combining statistical analysis, cohort segmentation, and dynamic visualization.

Project Overview

This dashboard simulates a full-fledged analytics workflow for an e-commerce company.
It provides end-to-end functionality from data ingestion and cleaning to KPI reporting, customer segmentation, retention analysis, and geographical sales intelligence.

The project is powered by a synthetic dataset (4,000+ orders across 2 years, 1,600+ customers, 10+ countries, and 5 categories).

Objectives

1. Tell a story with data: Convert large, unstructured datasets into interactive visual narratives.
2. Build an analyst-friendly interface: Enable filtering by country, channel, category, and time period.
3. Provide actionable insights: Identify best-performing channels, categories, and customer segments.
4. Demonstrate advanced analytics: Use RFM segmentation and cohort analysis to uncover retention patterns.

Key Metrics & Definitions

Analytics Features

KPI Cards | Summaries for Revenue, Profit, Orders, Customers, AOV, and Margin
Trend Charts | Monthly revenue & profit trends
Category & Product Insights | Top-performing product lines
Channel Revenue Share | Pie chart for sales by acquisition channel
Geographical Breakdown | Country → City treemap for global sales distribution
Cohort Retention Analysis | Track customer re-purchase behavior
RFM Segmentation | Classify customers into “Champions”, “Active”, and “New/Cold”

Dashboard Preview

KPI Overview & Monthly Revenue Trends

Monthly Revenue & Profit

Customers by RFM Segment

Channel Revenue Share

Revenue by Geography (Country → City)

Top 15 Products

Revenue by Category

Analytical Highlights

• Total Revenue: $9.8 M
• Total Profit: $3.0 M (~31% margin)
• Active Customers: 1,613
• Top Channel: Web (44.5%)
• Leading Category: Electronics (~ $2.3 M)
• Customer Segmentation:
  • 42% New/Cold
  • 30% Active
  • 28% Champions

These insights are based on synthetic two-year transaction data.

Tech Stack

Getting Started

Clone the Repository

git clone https://github.com/yourusername/data-storytelling-dashboard.git
cd data-storytelling-dashboard

Install Dependencies

pip install -r requirements.txt

Run the Dashboard

streamlit run app/app.py

The app will open at http://localhost:8501/.

(Optional) Use Your Own Dataset

export ORDERS_CSV=/path/to/your/orders.csv
# Windows PowerShell
$env:ORDERS_CSV="C:\path\orders.csv"

Your dataset must include:
order_id, order_date, customer_id, country, city, channel, product_id, category, subcategory, unit_price, quantity, discount, revenue, cost.

Folder Structure

data_storytelling_dashboard/
├── data/
│   └── orders.csv              # Synthetic e-commerce dataset
├── app/
│   ├── app.py                  # Streamlit main app
│   └── utils/
│       └── data_utils.py       # Functions for KPIs, filtering, cohort, RFM
├── requirements.txt
└── README.md

Possible Extensions

• Sales Forecasting (Prophet / ARIMA)
• Customer Churn Prediction using classification models
• Marketing ROI Analytics & A/B testing
• Enhanced Geo-maps using Plotly Choropleths
• Automated Insight Narratives (LLM-based summaries)
• Deployment via Streamlit Cloud / Render / Hugging Face Spaces

Insights Summary (Example Story)

“Between Jan 2023 and Oct 2024, overall revenue reached $9.8 M with an average margin of 30.9 %.
The Web and Mobile App channels contributed 75 % of total sales.
Electronics dominated category performance, while the India and Germany markets showed the highest growth rates.
Customer retention remains strong across cohorts, with ~30 % returning after 6 months.”