iNshiva/retail-sales-analysis-eswatini

Retail sales database built in PostgreSQL — 500 customers, 3000 orders, 9 SQL business queries with window functions and CTEs. Interactive dashboard built with Python and Plotly. # Retail Sales Analysis — Eswatini & Region (2021–2023) ### PostgreSQL · Python · Business Intelligence Dashboard --- ## Overview This project simulates a retail business database for a shop operating across Eswatini, South Africa, and Mozambique. Using PostgreSQL as the database engine and Python for analysis and visualisation, the project answers 9 real business questions that a retail manager would ask — from overall revenue performance to identifying the single most valuable customer for a loyalty award. The central goal: **demonstrate end-to-end SQL and business intelligence skills — from database design and data loading through to querying, analysis, and interactive dashboarding.** --- ## Key Business Findings - **Electronics is the top revenue category** at SZL 755,909 — driven by high-value items like Power Banks and Extension Cords - **Clothing has the best profit margin at 41%** — highest of all 8 categories - **Power Bank 10000mAh is the single best-selling product** at SZL 297,427 in revenue - **VIP customers (151 out of 491) generate 53% of total revenue** — the business is heavily dependent on its top tier - **Cape Town leads all regions** in both revenue and VIP customer count — cross-border sales are significant - **Lubombo VIP customers spend the most on average** (SZL 8,099 per customer) despite having fewer VIPs - **Loyalty Award Winner: Nompumelelo Mkhonta** from Hhohho, Eswatini — 9 orders, SZL 13,225 total spent since June 2022 --- ## Project Structure ``` retail-sales-analysis-eswatini/ │ ├── retail-sales-analysis-eswatini.py # Main analysis script ├── README.md # Project documentation ├── requirements.txt # Python dependencies │ └── visuals/ └── sales_dashboard.html # Interactive tabbed dashboard ``` --- ## Database Schema The project creates a normalised relational database with 6 tables: ``` regions └── customers (each customer belongs to a region) └── orders (eac …