From Data to Decision

Why this matters

Here is the file you will open in a few minutes: l01_retail_2025.csv. It holds one year of sales for Lantern & Pine, a home-goods retailer with 38 stores and an online channel — 50,539 rows, one row per product per order. Open it in a spreadsheet and you can scroll for twenty minutes without learning a single thing about the business.

By the end of this lesson, that same file will be a one-page report that answers, at a glance: how much did we sell, which categories carry the business, how did the year move, and how does any region compare to the others. The gap between those two experiences — 50,539 rows you cannot read, one page you can — is the entire reason business intelligence exists.

And the report will show you something the spreadsheet never would: the biggest sales month of 2025 was May, not December. May brought in $2,089,552; December, for all its holiday shopping, brought in $2,066,122. The reason is sitting in the category mix — Lantern & Pine sells fire pits and patio furniture, and outdoor gear peaks in early summer, hard enough to outweigh the holiday bump in everything else. Nobody finds that by scrolling. Everybody sees it the moment the line chart renders.

What business intelligence is

Business intelligence (BI) is the practice of turning an organization's data into information people can act on. Strip away the vendor language and it is a loop with five stations:

Today you will ride the whole loop once, fast, on purpose. The rest of the course slows down at each station. Keep this map in mind — every skill you learn for the next fourteen weeks lives at one of these five stations.

The Power BI ecosystem — three names you need to keep straight

Microsoft's naming here confuses everyone at first, so let us settle it on day one:

Getting set up

Install Power BI Desktop from the Microsoft Store (search "Power BI Desktop") or from powerbi.microsoft.com/desktop. The Store version updates itself, which is the behavior you want this semester. Desktop is free and does not require an account to build files — you will need your school account only when you publish in the last section.

Then download two course files: l01_retail_2025.csv (the data) and iris-theme.json (the course report theme — a set of colors and defaults chosen to be readable and colorblind-safe; Module 3 explains why that matters).

First contact: three views

When Desktop opens you land in Report view — a blank canvas, a Visualizations pane, a Data pane. Two more icons sit under it on the left rail. Table view shows the rows of whichever table you select, like a spreadsheet you cannot type into. Model view shows your tables as boxes and the relationships between them as lines — with one table it looks empty today, and in Module 2 it becomes the most important screen in the application.

One habit to build immediately: the canvas is not the data. A visual on the canvas is a question you have asked; Table view is where you go to see what the data actually says row by row. When a number looks wrong — and this course will show you many numbers that look wrong — Table view is where you check.

Load the data

On the Home ribbon, choose Get data → Text/CSV, select l01_retail_2025.csv, and Desktop shows you a preview: seven columns — OrderID, OrderDate, Region, Category, Product, Quantity, Revenue — and the first rows of the data.

The preview dialog offers two buttons: Load and Transform Data. Transform Data opens Power Query, the cleaning workshop where Lessons 2 and 3 live. Today the file is clean — it was prepared so that nothing stands between you and your first report — so click Load. In Lesson 2 you will learn how rarely real files earn that click.

RevenueThe money a business takes in from selling products or services, before any costs are subtracted. in this file is already computed per row: quantity × unit price, minus any discount. Later in the course you will compute it yourself from the raw pieces; today it is one less thing in the way.

Build the report — four visuals, one page

1 · The headline number: a card

In the Data pane, check Revenue. Power BI drops a column chart on the canvas — its default guess. With the visual selected, choose the Card icon in the Visualizations pane. The chart collapses to a single number: $20.66 million — precisely $20,663,252.72, the sum of the Revenue column across all 50,539 rows. That is the year, in one figure. Notice what you did not do: you never wrote a formula. Power BI saw a numeric column and summed it. That automatic guess is called an implicit measure, and Module 2 will show you both its convenience and the trouble it can cause.

2 · What carries the business: a bar chart

Click an empty part of the canvas, then check Revenue and Category together. You get a bar chart, and it already tells a story: Furniture leads at $7,592,351, Outdoor follows at $5,553,569, and Decor trails the field at $1,612,488. Two categories are carrying more than half the company's revenue.

3 · How the year moved: a line chart

Empty canvas again; check Revenue and OrderDate, and switch the visual to a line chart. Power BI first shows one point per year — it groups dates into a hierarchy of Year, Quarter, Month, Day. Click the forked "expand" arrow in the visual's header until you see months. There is the shape of the year: a February trough of $1,142,332, the climb to that surprising May peak of $2,089,552, and December's $2,066,122 falling just short of it.

4 · Let the reader ask questions: a slicer

One more: check Region, and change the visual to a slicer. A slicer is a filter that lives on the page, owned by whoever is reading the report. Click Online and watch every visual on the page recalculate at once: $9,218,459 of revenue — about 45 cents of every dollar — never touches a store. Click Southwest and the same page tells a different, smaller story: $1,758,462, the quietest region in the company.

That live recalculation is called cross-filtering, and it is the single biggest thing a Power BI report can do that a printed chart cannot: the reader interrogates the page instead of reading it.

Publish it

On the Home ribbon, choose Publish, sign in with your school account, and pick My workspace. Desktop uploads the report to the Power BI Service; the success dialog gives you a link. Open it: your report, in a browser, slicers working, no Desktop required. Anyone you share that link with (Module 4 covers who can see what) reads the live report — not a screenshot, not an email attachment with final_v3_REAL.pbix in the filename.

Where BI shows up — three industries, same loop

• Retail (this course's home turf). Category mix, seasonality, store-versus-online — the exact questions you answered today, asked weekly by every merchandising team on the planet.
• Health care. Patient volumes by clinic and month, wait times, readmission rates. The line chart with the surprising peak is a flu season instead of a fire-pit season; the loop is identical.
• Finance. Branch performance, loan portfolios by risk tier, month-end close packs that used to take a week of copy-paste and now refresh themselves. Module 4's scheduled refresh is why finance teams adopted BI first.

Common mistakes

• Skipping the type check in the preview dialog. If OrderDate loads as text, the line chart cannot group by month and sorts "April" before "February" — alphabetically. Nothing errors. The chart just quietly stops meaning anything.
• Treating the canvas as the data. When a number surprises you, go to Table view and look at rows. The visual is an answer to a question; make sure it was your question.
• Confusing Desktop and the Service. If a classmate "can't find Power Query in the browser," they are in the Service. Building happens in Desktop.
• Forgetting which file is open. Desktop's title bar shows the file name. Assignments in this course are submitted as .pbix files with required names — starting a new blank file and rebuilding is not the same as saving the one specified.

Summary