Reporting and KPIs
Building Power BI Reports: A Step-by-Step Guide for Beginners
How to build a Power BI report step by step: connect data, shape it in Power Query, model relationships, design visuals, and publish. A beginner walkthrough.
Sajagan Thirugnanam
·
Updated
Building Power BI reports means connecting to a data source, shaping the data in Power Query, modeling the relationships between tables, then adding and formatting visuals before you publish. Power BI Desktop is free to download and is where you do steps 1 through 6 of building a Power BI report; publishing in step 7 needs a workspace and, in most cases, a Pro or PPU license.
How to Create a Power BI Report, Step by Step
The seven steps below take you from a blank canvas to a published report.
Step 1: Install Power BI Desktop
Download Power BI Desktop from the Microsoft site or from the Microsoft Store. Both give you the same application. Once it's installed, open it. You start on a blank canvas, ready to connect to data.
Step 2: Connect to a data source
Power BI Desktop's Get Data dialog, listing the available connector categories.
Power BI connects to Excel files, SQL Server, SharePoint, Azure services, Salesforce and many other sources.
Select Get Data on the Home ribbon.
Choose the data source from the list.
Enter the connection details: a file location for Excel, or a server name and credentials for a database.
Preview the data and select Load.
Step 3: Shape the data in Power Query
After you connect, Power BI shows a preview you can edit in Power Query Editor before it's added to the model. Common cleanup steps:
Remove duplicate rows and columns you don't need.
Rename columns and fix data types, especially dates and numbers imported as text.
Split or merge columns to get the structure you need.
Filter out rows that don't belong in the report, and apply any calculations that should happen once, at load, rather than in every visual.
This step decides how reliable everything downstream is. A column with the wrong data type or a duplicate row here shows up as a wrong number in a visual later.
Step 4: Build the data model
Model view, showing relationship lines between a fact table and its dimension tables.
If your report uses more than one table, you need relationships between them.
Switch to Model view.
Drag a field from one table onto the matching field in another to create a relationship.
Where you can, use a star schema: one fact table (such as Sales) connected to dimension tables (such as Customers, Products and Dates). A star schema keeps relationships simple and measures fast. See our guide to star schema versus snowflake schema for why this matters for performance.
Step 5: Design the report
A report page with a bar chart, a table and a card visual placed on the canvas.
Pick a visual: bar chart, line chart, table, matrix, map or KPI card, based on what the question needs.
Drag fields into the visual's field wells: Values, Axis, Legend or Tooltip, depending on the visual.
Add filters and slicers. Use the Filters pane for filters that apply across the page or report, and slicers on the canvas for filters the reader controls. Our guide to Power BI filters covers the difference in more detail.
Most numbers a reader wants (a total, an average, a year-over-year change) should be a DAX measure (a formula that recalculates as filters change) rather than a raw column, which stays fixed. See our complete DAX guide if you haven't written a measure yet.
Step 6: Format the report
The Format pane, open for a selected visual.
Use consistent colors and fonts, ideally from a saved theme, across every page.
Put the most important numbers at the top or top-left, since that's where readers look first.
Leave whitespace. A crowded page is harder to read than a sparse one.
Add titles and tooltips so a reader doesn't have to guess what a visual shows.
Use the Format pane, opened from the paint-roller icon, to adjust fonts, backgrounds and gridlines on the selected visual.
Add bookmarks and buttons if the report needs guided navigation between views.
Step 7: Publish and share
Once the report is ready:
Save it in Power BI Desktop.
Select Publish on the Home tab.
Choose the workspace to publish to.
From there you have three ways to share it:
Direct share, granting access to specific people.
Workspace access, for a team that collaborates on the report itself. See our guide to workspaces and apps for the difference between adding someone to a workspace and giving them an app.
An app, which packages one or more reports for a wider audience without giving them access to the workspace's editing surface.
Anyone who edits, comments on, or shares the report from the service needs a Power BI Pro or PPU license unless the workspace is on Premium capacity. See our guide to Power BI licensing for current pricing and what each license tier includes.
Licensing and limits to know
Power BI Desktop is free. Everything through step 6 needs no license.
Pro and PPU licenses are priced per user, per month, billed yearly: Pro at $14.00 and PPU at $24.00. You need one to publish, share, or collaborate on a report in the Power BI service. Readers without a license can open a shared report only when its workspace sits on a Fabric F64 or larger capacity, or a Power BI Premium P capacity.
Premium capacity has moved to Fabric F SKUs. Power BI Premium P-SKU subscriptions are being retired; new capacity purchases are Fabric F SKUs instead.
Free-license viewers need a large enough capacity. A user with no Pro or PPU license can only view Power BI content when the workspace is on an F64 or larger Fabric capacity, or a legacy Premium P SKU. On a smaller capacity, they still need a license.
A workspace can hold up to 1,000 semantic models and reports. Large enough for most teams, but worth knowing if you're consolidating many reports into one place.
Exporting and integration
Export a report to PDF or PowerPoint for static sharing, or export the underlying data to Excel or CSV for further analysis.
Use Analyze in Excel to work with the semantic model's data directly in Excel.
Embed a report in Microsoft Teams, SharePoint or a website using Power BI's embed options.
Sync a report's files through OneDrive or SharePoint for version history.
What to learn next
Once the basics work, the next things worth learning are DAX time intelligence, drill-through pages, and how row-level security restricts what different readers see in the same report.
FAQs
Do I need coding skills to use Power BI?
No. Most of report building is drag-and-drop: connecting data, adding visuals, and formatting them. Learning DAX lets you write your own calculations instead of relying only on the columns already in your data, but it's not required to build a working report.
What's the best data model for Power BI?
A star schema, with one fact table connected to several dimension tables, is the model Microsoft recommends. It keeps relationships simple, avoids ambiguous filter paths, and gives DAX measures a clear, predictable table to filter against.
Do I need a Power BI Pro license to build a report?
No. Power BI Desktop, where you build and shape a report, is free. You need a Pro or PPU license to publish it to a workspace and share it. The people you share it with also need Pro or PPU, unless the workspace is on a Fabric F64 or larger capacity or a Power BI Premium P capacity.
How long does it take to build a first Power BI report?
It depends on the data. A single clean table with a few visuals can take under an hour. A report with several related tables, a proper star schema and custom DAX measures usually takes longer, because most of the time goes into shaping data correctly in steps 2 through 4, not into placing visuals in step 5.
Sources
What is Power BI Premium? - Microsoft Learn
Workspaces in Power BI - Microsoft Learn
