Twelve weeks can give you a structured Power BI practice cycle. It cannot guarantee mastery, certification or a job. The target is one defensible case study and enough coverage of the current data-analyst role to know what you need to practise next.
Use the PL-300 study guide as a current skills map, even if you do not plan to take the exam.
Weeks 1 and 2: define the question and inspect the data
Choose a problem from a domain you understand. Write the decision, audience and success measure before opening Power BI.
Import a safe public or synthetic data set. Use Power Query to profile columns, assign types, handle errors and document nulls. Keep a short data dictionary. Record what the source cannot tell you.
Evidence to keep: the question, source note, data dictionary and transformation decisions.
Weeks 3 and 4: build the model
Separate facts from dimensions. Create explicit relationships and a proper date table. Prefer a star schema unless there is a justified reason not to.
Create a small set of explicit measures. Test each against a simple table before using it in a polished visual. Explain filter context in your own words.
Evidence to keep: model diagram, relationship notes and measure checks.
Weeks 5 and 6: design the report
Match visuals to questions. Add useful slicers, titles and tooltips. Test interactions, drillthrough and reset behaviour. Check keyboard order, contrast and alt text.
Ask somebody unfamiliar with the project to explain what they think the report says. Their confusion is a design finding, not a user failure.
Evidence to keep: before-and-after screenshots and a record of the feedback you acted on.
Weeks 7 and 8: use the Power BI service
Publish to an appropriate non-production workspace. Learn how semantic models, reports, workspaces and apps differ. Configure credentials and refresh only after checking the source and gateway requirements.
Understand workspace roles and row-level security. Test as a viewer, not only as the owner. Check current licence requirements for the intended sharing route.
Evidence to keep: refresh history, access test and documented owner.
Weeks 9 and 10: diagnose and improve
Use Performance Analyzer to find slow report interactions. Inspect model size, unnecessary columns, expensive measures and source behaviour. Change one thing at a time and record the evidence.
Add a reconciliation page or external check for the measures that matter. A beautiful total that cannot be traced back to the source is not ready.
Evidence to keep: test cases, performance observations and reconciliation results.
Weeks 11 and 12: package the case study
Write a short portfolio narrative:
- The business question.
- The approved data source and its limitations.
- The model and measure decisions.
- The report choices.
- The refresh and security design.
- What you tested.
- What remains outside the proof boundary.
Do not publish confidential files or imply a business outcome you did not measure. If the work uses sample data, say so.
Sources
- Study guide for Exam PL-300
- Understand star schema in Power BI
- Power BI implementation planning
- Create accessible Power BI reports
Keep the plan grounded in real reporting work
The Microsoft 365 Report Builders Space focuses on practical reporting workflows, reliable metrics and the Power BI decisions that survive beyond a polished demo.