At some point, someone in your organisation will say the words “can we get this in Power BI?” — usually while pointing at a perfectly good spreadsheet. Before you nod and open forty browser tabs, here’s the thing the tutorials don’t lead with: if you’ve climbed the Excel path, you already know half of Power BI. It’s made by the same company, out of the same parts, for the same job — and the parts transfer.
This is the honest orientation: what Power BI is, what carries over, what’s genuinely new, and a first project sized for one afternoon.
What it actually is
Power BI Desktop is a free Windows application that does three things in one window:
- Gets and cleans data — using, quite literally, Power Query. Not a lookalike: the same engine, same editor, same Applied Steps. If you’ve combined a folder of monthly files in Excel, you already hold a Power BI skill.
- Relates tables — more on this below; it’s the real new idea.
- Builds interactive report pages — where PivotTable thinking (fields dragged into wells, data aggregated on demand) becomes charts, maps and slicers that all filter each other.
The division of labour in one line: Excel is where you work a
model; Power BI is where a report goes to be consumed — live,
clickable, on phones, without anyone emailing
Dashboard_FINAL_v7.xlsx ever again.
The one genuinely new idea: the model
In Excel, when data lives in two tables, you reach for XLOOKUP and physically drag columns across. Power BI says: don’t. Load both tables, draw a relationship between them (customer ID here equals customer ID there), and every visual can now use fields from both — no lookup columns, no duplication.
That mental shift — relate, don’t merge — is the actual learning
curve of Power BI. It usually takes an Excel person a weekend, and
it’s the same idea that later unlocks proper data modelling
anywhere. (You’ll also meet DAX, the formula language, guarding
this stage like a customs officer. Opening position: write measures
when a card needs a number — Total Sales = SUM(Sales[Value]) reads
like a sentence — and resist copying clever DAX from forums until a
real report demands it.)
Your first project, honestly sized
Pick a dataset you already understand — your budget transactions are ideal, because you know what the answers should look like. Then, in one afternoon:
- Load your transactions Table from the workbook (
Get Data → Excel). Do any cleaning in the query editor — familiar territory, same recipe thinking. - Build four visuals: a card showing total spend, a bar chart by category (sorted — the chart rules apply verbatim here), a line by month, and a slicer for the month. No relationships needed yet; one table is fine for day one.
- Click things. Click a category bar and watch every other visual filter to match. This cross-filtering is the moment Power BI justifies itself — the interactivity you can’t build with a static chart.
That’s a real dashboard, built with roughly 80% recycled Excel knowledge.
What they don’t tell you in the demo
Fair warnings from the field. Sharing is where the free ride ends — Desktop costs nothing, but publishing dashboards to colleagues needs Pro licences (bundled with some Microsoft 365 plans, ~£8/user/month otherwise). Refresh needs plumbing: a published report doesn’t magically see your laptop’s files. And plenty of jobs should stay in Excel — ad-hoc analysis, anything with manual inputs and judgement, models you interrogate cell by cell. The failure mode of Power BI enthusiasm is rebuilding a working spreadsheet as a slower dashboard nobody filters.
The right first move is the afternoon project above: cheap, honest, and it tells you whether your reporting problems are dashboard-shaped. Many are; some never will be.
You spent the whole Excel path learning to structure data, aggregate it honestly, and chart it truthfully. Power BI is those same lessons with a publish button. Go click a bar chart and watch the room filter — you’ve earned it.