Module 1: Getting started · Lesson 1.2 · about 12 min
Meet Northbay Goods: the data for this course
A made-up Ontario retail chain with five stores, twelve products and two years of order lines. Download the five CSV files, see how the tables fit together, and learn what one row of each table means.
Module 1 · about 12 minutes · read first: 1.1. You’ll be able to download the course data, describe each table’s grain, and read the model diagram.
1. Why this matters
Learning on real-looking data beats learning on a list of fruit. Northbay Goods is a made-up chain, but it behaves like a real retailer: weekends are busier, snow shovels sell in winter, earbuds peak before the holidays, prices rose about 4 % in 2025 and a fifth store opened in March 2025. Those quirks are on purpose: each one turns into a lesson (a store that did not exist last year breaks naive growth measures, for example).
2. The model in one picture
3. The five tables
| Table | One row is… | Key columns |
|---|---|---|
| Sales (fact) | one order line: a product bought in a store on a day | Date, StoreKey, ProductKey, Quantity, UnitPrice, UnitCost, Amount |
| Product | one product | ProductKey, Product, Category |
| Store | one store | StoreKey, Store, City, Region, OpenDate |
| Date | one calendar day, 2024 to 2025 | Date, Year, Quarter, Month, YearMonth, Weekday |
| Targets | one store’s sales target for one month | YearMonth, StoreKey, Target |
Fact tables record events and hold numbers you add up (Quantity, Amount). Dimension tables describe the who, what, where and when, and hold the words you filter and group by (Category, Region, Month). Module 3 goes deeper.
Amount is Quantity × UnitPrice, already worked out, as many till systems do. UnitCost lets you work out margin.
4. Get the files
Five CSV files, made right here in your browser from a fixed seed, so your numbers match the lessons exactly.
Preparing the tables…
You don’t need the files for the practice editors in this course: they run on the same data inside your browser. You need them for the hands-on parts in Power BI Desktop, starting in Lesson 1.4.
Principle
Principle: before any formula, know the grain of each table: what one row means. Most wrong numbers in BI come from adding up rows at the wrong grain (for example, adding a monthly target to every day).
Rule: a fact table’s grain is the most detailed event it records. Every column in it must be true for exactly that event.
Try it: what is the grain of a table with columns StoreKey, Week, Staff Hours? (One store for one week.)
5. Worked example
How many order lines did Queen Street have on its busiest day? You’d filter Store to Queen Street, group Sales by Date, count rows per date and take the largest. In DAX that becomes MAXX(VALUES('Date'[Date]), CALCULATE(COUNTROWS(Sales))), which you’ll be able to read by the end of Module 5. For now, notice the plan: filter a dimension, count fact rows, group by another dimension.
6. Practice
- Which table would you filter to see only the East region, and which table holds the numbers that change?
- Store 5 (Forks Market) opened on 2025-03-01. What will its sales be for January 2024?
- Is “Category” a fact or a dimension attribute?
Answers
- Filter Store[Region]; the numbers are in Sales.
- Nothing (blank): it had no order lines before it opened.
- A dimension attribute: it describes a product and is used to group and filter.
Check yourself
Your ticks stay in this browser. Save them as a file on the course home.
Updated 2026-11-22. Northbay Goods is a made-up shop and its numbers are generated. Power BI changes every month: menu names can move, so check Microsoft Learn if a button is not where the lesson says. The practice editor runs in your browser; nothing you type is sent anywhere.