Northwind Dashboard

This repository contains a start-to-finish overview of building a sales dashboard for stakeholders at a fictional ‘Northwind’ shipping company.

It combines PowerBI, PostgreSQL, GitBash and R to create succinct insights on the company's main KPIs (AOV, revenue) as well as dedicated tabs for specific categories.

Step 1: Setup Postgres database

This GitHub repo contains the raw data for hosting Northwind on a PostgresSQL database. Using GitBash and PgAdmin, I could extract, transform and load the data into a locally hosted database.

Step 2: Data transformation and analysis

Once inside the database, I used SQL scripts to join the data based on the Northwind schema and transform the data and used CTEs and window functions to explore top-level trends like ranking countries by revenue or calculating average order value.

SQL example - Rank countries by revenue

WITH revenue_tab AS(
SELECT 
    ship_country,
    ROUND(SUM((unit_price * quantity * (1 - discount))::numeric),2) AS total_revenue
FROM orders o
JOIN order_details od ON o.order_id = od.order_id
GROUP BY ship_country)


SELECT 
    ship_country, 
    total_revenue,
    DENSE_RANK() OVER(
       ORDER BY total_revenue DESC) AS rank
FROM revenue_tab;

SQL example - Calculate average order value

SELECT 
    ROUND(AVG(total_revenuel), 2) AS average_order_value
FROM order_revenue;

Step 3: Dashboard development

I connected PowerBI to the database and used DAX expressions to write measures for critical KPIs like average order value (AOV) and YoY change in revenue, profits and orders.

Step 4: Dashboard design

Design focuses on delivering clear, summary insights at a glance. It included:

• Clear data visualizations
• A succinct summary page
• Categories for key business branches (sales, inventory, customers)
• Developing a colour palette and logo
• Building custom visualizations in R

Screenshots