The short version
This is a review sheet, not a course. Use it in the last days before the exam to check that the main distinctions are clear in your head. It does not cover everything on the exam, and it contains no real exam questions. The facts follow Microsoft’s study guide “Skills measured as of April 20, 2026”, and the rules and explanations are our own wording. If you cannot explain a line out loud, go back to domains explained and practice it in Power BI.
Exam facts at a glance
| Item | Fact (checked October 22, 2026) |
|---|---|
| Exam | PL-300: Microsoft Power BI Data Analyst |
| Time | 100 minutes |
| Pass mark | 700 on a scale of 1 to 1,000 (scaled, not 70 percent) |
| Questions | Number not published; interactive components possible |
| Delivery | Proctored, scheduled through Pearson VUE |
| Retake | 24 hours after the first failure, then 14 days; five tries in 12 months |
| Credential | Valid for 12 months; free online renewal |
| Prepare the data | 25-30% |
| Model the data | 25-30% |
| Visualize and analyze the data | 25-30% |
| Manage and secure Power BI | 15-20% |
No penalty for wrong answers, so answer everything. Some items are unscored, so treat each one as if it counts.
Prepare the data
Key concepts
- Connection modes: Import copies data into the model, DirectQuery sends queries to the source, and Direct Lake reads data from Fabric storage.
- Data source settings hold credentials and privacy levels.
- Parameters make a source or filter value changeable.
- Profiling shows quality, distribution, and statistics for columns.
- Transformations: group, pivot, unpivot, transpose, merge, append, and convert semi-structured data (such as JSON) to a table.
- Fact tables hold events and numbers. Dimension tables hold descriptions. Relationships need keys.
Easily confused pairs
- Merge adds columns by matching rows. Append stacks rows.
- Reference builds on the output of another query, so changes upstream flow through. Duplicate is an independent copy of the steps.
- Pivot turns rows into columns. Unpivot turns columns into rows (use it when months are columns).
- Import is fast and has a refresh. DirectQuery is fresh and depends on the source.
- Replace values changes cell content. Remove empty or Remove errors removes rows. Know which one a scenario needs.
Decision rules
- If columns are values (Jan, Feb, Mar), think unpivot.
- If you must keep every row of one table and add matches from another, think left outer merge. If you must find rows with no match, think left anti.
- If the scenario asks for the smallest model, remove unneeded rows and columns before load.
- If a source needs a changing value such as an environment name, think parameter.
Traps
- A wrong data type early in the chain breaks later steps.
- A disabled load setting means the query exists but the table does not appear in the model.
Model the data
Key concepts
- Star schema: one fact table in the middle, dimensions around it, one-to-many relationships.
- Cardinality and cross-filter direction decide how filters flow.
- A common date table supports time intelligence and should be marked as a date table.
- Role-playing dimension: one dimension used for several roles (order date, ship date) with one active relationship and the others inactive.
- DAX: measures react to filter context, calculated columns are computed at refresh, CALCULATE changes filter context, and semi-additive measures (such as stock balance) do not sum across time.
- Calculation groups reuse logic such as time intelligence across many measures.
- Performance Analyzer shows where a visual spends time. DAX query view lets you run and inspect queries.
Easily confused pairs
- Calculated column is row by row, stored in the model. Measure is aggregated on the fly, not stored.
- SUM adds a column. SUMX goes row by row, then adds.
- Single direction filtering is the default and the safest. Both directions can fix a result and add ambiguity or slowdown.
- Inactive relationship needs a measure that activates it. Active relationship works automatically.
- Calculated table is created in the model. A Power Query table is created before load.
Decision rules
- If the number must change with slicers, think measure.
- If you need to slice or filter by a new value per row, think column.
- If dates are missing or inconsistent, build one date table and relate the facts to it.
- If a question says “year to date” or “same period last year”, think time intelligence on a proper date table.
- If a model is slow, remove unused columns, reduce granularity, and look at heavy measures.
Traps
- Reading CALCULATE as “add a filter” when the filter replaces an existing filter on the same column.
- Setting both-way filters everywhere to avoid thinking about the model.
Visualize and analyze the data
Key concepts
- Match the visual to the question: comparison, trend, composition, distribution, ranking.
- Themes set consistent colors and fonts. Conditional formatting colors by rule or value.
- Slicers and filters limit data. Sync slicers apply one choice across pages.
- Bookmarks save a view. Drillthrough sends a user to a detail page with the filter. Tooltips show detail on hover.
- Visual calculations are DAX written on the visual itself.
- Accessibility: alt text, tab order, color contrast, and readable titles.
- Patterns: grouping, binning, clustering, forecasting, reference lines, error bars, anomaly detection, AI visuals.
- Paginated reports fit exact, printable layouts.
Easily confused pairs
- Drill down moves down a hierarchy in the same visual. Drillthrough opens another page.
- Filter removes other data. Highlight dims it and keeps context.
- Layer order (what sits on top) and tab order (keyboard sequence) are separate lists in the Selection pane.
- Dashboard (a service canvas of pinned tiles) is not a report (pages of visuals).
Decision rules
- If the requirement is a pixel-perfect printed layout, think paginated.
- If a user needs detail without leaving the page, think tooltip. If they need a full detail page, think drillthrough.
- If keyboard users or screen readers are mentioned, think tab order, alt text, and contrast.
- If a report must update itself, think automatic page refresh.
Traps
- Choosing a pretty visual that hides the comparison the question asks for.
- Forgetting that a mobile layout is a separate design for the same report.
Manage and secure Power BI
Key concepts
- A workspace holds items. An app is a packaged, curated view for an audience.
- Distribution methods differ in who gets access and how much they can do.
- Subscriptions send report snapshots. Data alerts fire when a value crosses a threshold.
- Endorsement: promote or certify content to signal trust.
- A gateway connects the service to on-premises data. Scheduled refresh updates a model.
- Workspace roles are broad. Item-level access is narrow.
- Row-level security (RLS) filters rows for roles. Sensitivity labels classify and protect content.
Easily confused pairs
- Promote and certify are both endorsement, but they signal different levels of trust and different people may apply them. Know the difference in Power BI before exam day.
- Workspace role gives access to everything in the workspace. Item-level access gives access to one item.
- RLS limits rows inside a model. It does not grant access to the report by itself.
- Personal gateway versus standard gateway: shared use and scope differ.
Decision rules
- If an on-premises source needs scheduled refresh, think gateway.
- If each manager must see only their region, think RLS plus a way to assign users or groups.
- If a person needs only to view one report, think item-level access or an app, not a high workspace role.
- If data must be marked as confidential, think sensitivity label.
Traps
- Granting a high workspace role to solve a small access need.
- Assuming a user in an RLS role can open the content without separate access.
Time and tactics
- Answer in the first pass, mark doubts, and review at the end. Microsoft says that once you take a break you cannot go back to earlier questions, so finish your review first.
- Read the last sentence of a scenario first, then the whole thing. Words such as “must”, “only”, and “minimum” decide the answer.
- Roughly two minutes per question is a working rule, not a vendor number.
Last-day checklist
- I can explain Import, DirectQuery, and Direct Lake in one sentence each.
- I can say what merge, append, Reference, and Duplicate do.
- I can draw a star schema and set filter direction.
- I can explain a measure versus a calculated column and what CALCULATE does.
- I can choose a visual and tell drill down from drillthrough.
- I know when a gateway is needed.
- I can describe RLS, workspace roles, and item-level access.
- I know my exam time, ID rules, and room or online setup (see the exam day guide).
- I slept, and I stopped learning new topics.
Power BI features change often, so check the current Microsoft study guide for the skill list. It is the final authority.
More guides for PL-300
- PL-300 Exam Format, Scoring and Retakes
- PL-300 Study Plan: 4-Week and 8-Week
- Is PL-300 Hard? An Honest Difficulty Check
- PL-300 Prerequisites and Skills Check
- PL-300 Renewal and Retakes Explained
- Is PL-300 Worth It? An Honest Look
- DP-600 vs PL-300: Which Exam First?
Next step
Put these lines to the test with the sample questions, and revisit domains explained for the weak ones. When you want a full run, our PL-300 practice tests are original practice sets by domain, back to back with the exam overview.