On this page
Intro
PL-300, Microsoft Power BI Data Analyst, supports the Microsoft Certified: Power BI Data Analyst Associate credential. Microsoft describes the candidate as someone who delivers actionable insights from available data, builds easy-to-understand visualizations and enables others to do self-service analytics. The exam expects you to be proficient with Power Query and Data Analysis Expressions (DAX). CertMonk can help organize review alongside practical work in Power BI Desktop. The audience description comes from Microsoft’s certification page.
Because the exam is built around a workflow, from connecting to data to securing a published report, the most effective preparation follows that same path with a small dataset of your own.
Exam at a glance
| Detail | PL-300 |
|---|---|
| Exam code | PL-300 |
| Questions | Exact count not published on the reviewed pages; Microsoft says question numbers vary by exam |
| Time | 100 minutes, as listed on the certification page |
| Passing score | 700 or greater on a 1–1,000 scale |
| Question types | Proctored exam; interactive components may appear |
| Price | Check the vendor’s site for current pricing |
| Retirement | Date not published on the reviewed pages |
Sources: PL-300 certification page, exam experience and scoring policy, reviewed November 7, 2026.
Version update. Microsoft’s study guide lists skills measured as of April 20, 2026, with minor changes to getting data, report usability and workspace management. Copilot-related report tasks now appear in the outline. Check any course or book against the current study guide before relying on it.
Exam domains and weights
| Domain | Weight | In plain English |
|---|---|---|
| Prepare the data | 25–30% | Connect to sources, clean and shape data, and load it. |
| Model the data | 25–30% | Design relationships, write DAX measures and tune performance. |
| Visualize and analyze the data | 25–30% | Build reports, improve usability and find patterns. |
| Manage and secure Power BI | 15–20% | Organize workspaces, share content and control access. |
Microsoft publishes ranges. Three of the four domains carry the same range, so an even study split is a sensible starting point, with extra time on whichever area your practice results flag. Use the skills measured list as your checklist.
What each domain really tests
Prepare the data
Study this domain as a decision sequence. First choose how to connect: the outline asks you to choose between DirectLake, DirectQuery and Import, so compare their trade-offs in your own words. Then profile the data, resolve nulls and inconsistencies, and decide how to handle import errors.
Transformation is the larger share. Practice grouping and aggregating rows, pivoting and unpivoting, merging and appending queries, and converting semi-structured data into tables. For each transformation, write what the data looked like before and after, and why the step belongs in Power Query rather than in the model.
Pay attention to the idea of fact and dimension tables, to keys that support relationships and to the difference between reference and duplicate queries. These choices shape everything that follows.
Model the data
This domain tests whether your model behaves correctly. Start with structure: table and column properties, relationship cardinality, cross-filter direction, role-playing dimensions and a common date table. Draw the model before you build it.
For DAX, build understanding from simple aggregation measures to the CALCULATE function, then to time intelligence, semi-additive measures, calculation groups and statistical functions. Explain filter context aloud for each measure: which filters apply, which are removed and which are added.
Performance is part of the outline. Practice removing unnecessary rows and columns, reducing granularity and using Performance Analyzer and DAX query view to find slow measures, relationships or visuals. Record what you changed and what the evidence showed afterward.
Visualize and analyze the data
Choose visuals from the question being asked, not from habit. For each report page, write the question it answers, then pick a visual, configure formatting and apply filtering and slicing. The outline also lists themes, conditional formatting, paginated report decisions and visual calculations.
Usability topics include bookmarks, custom tooltips, visual interactions, navigation, sync slicers, drillthrough, export settings, mobile layouts, personalization, accessibility and automatic page refresh. Build one report that uses several of these together so you see how they interact.
For analysis, practice the Analyze feature, grouping and binning, clustering, AI visuals, reference lines, error bars, forecasting and anomaly detection. The current outline includes Copilot tasks such as creating a narrative visual and summarizing a semantic model; give them their own notes.
Manage and secure Power BI
Trace content from workspace to audience. Learn how to create and configure a workspace, publish or update items, configure an app, choose a distribution method, set subscriptions and data alerts, and promote or certify content. Know when a gateway is required and how to configure a scheduled refresh.
Security topics include workspace roles, item-level access, access to semantic models, row-level security roles with group membership and sensitivity labels. For each, write who can see what before and after the change. A diagram of roles against content is more useful than a list of definitions.
The topics above follow Microsoft’s PL-300 outline; the suggested exercises are our learning approach.
A realistic six-week study plan
This plan assumes you have opened Power BI Desktop before. It is a suggested schedule, not a promise about readiness. Add time if basic spreadsheet or data concepts are new to you.
Week 1: connect and profile
Read the outline once and pick a practice dataset. Connect to it, compare connection modes and profile the data. Write down every quality problem you find.
End the week by explaining why you chose your connection mode and what you would change for a much larger dataset.
Week 2: transform and load
Work through column types, grouping, pivoting, merging, appending and keys. Keep a log of each applied step and its purpose.
Add untimed practice and note the reasoning behind each uncertain answer.
Week 3: build the model
Design relationships, a date table and role-playing dimensions. Set properties and test filter behavior with a few simple visuals.
Leave one session for catch-up. When a relationship behaves unexpectedly, trace the filter path step by step before changing anything.
Week 4: DAX and performance
Write measures from simple to advanced, including time intelligence and calculation groups. Explain filter context for each. Then use Performance Analyzer on your own report and make one evidence-based improvement.
Week 5: reports, storytelling and analysis
Build a multi-page report with bookmarks, drillthrough, tooltips and accessibility settings. Add analysis features and review the Copilot tasks in the outline.
Use mixed practice to find weak links between modeling and visuals.
Week 6: management, security and review
Publish your report to a workspace, configure refresh and apply row-level security. Then add timed practice and inspect your reasoning afterward.
Make a deliberate readiness decision. If your notes contain many unexplained steps, extend the schedule with specific goals. Finishing six weeks is a milestone, not proof of readiness.
Common mistakes and how to avoid them
- Skipping DAX fundamentals. Learn filter context before memorizing function names.
- Cleaning data in the wrong place. Decide whether a change belongs in Power Query, in the model or in a measure.
- Ignoring relationship direction. Test filters with small visuals before building the full report.
- Treating the management domain as minor. It carries a published range, and security questions need careful reading.
- Using an older outline. The skills measured were updated in April 2026, so compare older material with the current guide.
- Rushing into timed practice. Build clear explanations untimed first, then use the clock to check pacing.
How to use practice questions well
After each practice item, explain why the chosen answer fits and why each alternative does not. Review wrong-answer explanations for the assumption that misled you, and add that concept to your next session.
CertMonk includes explanations for every answer choice, randomized practice tests, timed and untimed modes, flip-card flashcards, study games and progress by exam domain. Use the domain view to choose follow-up work, and flashcards for brief review between hands-on sessions.
Keep a separate list of correct guesses and revisit them later without the answer choices in front of you. When you add timed practice, use the result to investigate your preparation rather than to predict an exam outcome.
Free practice questions
Use the free PL-300 samples below to explore the explanations and spot areas to review without a purchase.
A CSV file from a European system uses semicolons between fields, and accented names such as "Müller" appear garbled after import. Which two settings of the Text/CSV connector address these problems? (Choose two.)
Choose 2.
- AFile origin, set to the encoding the file was saved with
- BDelimiter, set to Semicolon
- CColumn profiling on the entire data set
- DData type detection, set to not detect types
- EFuzzy matching on the Name column
Show the answer and explanation
Answer: A. File origin, set to the encoding the file was saved with and B. Delimiter, set to Semicolon
Correct. File origin tells the connector how the file’s characters are encoded, which fixes garbled accented characters.
Correct. The connector lets you specify the delimiter, so semicolon-separated fields split into columns.
Why not the others
- C. Profiling reports on values but does not change how the file is read.
- D. Turning off detection leaves every column as text; it does not split the fields or fix the characters.
- E. Fuzzy matching is a merge option; it does not decode characters in a file.
Takeaway: For text files, set File origin to fix character encoding and choose the right delimiter so fields split correctly.
Source: Text/CSV - Power Query connector
A report page takes roughly 20 seconds to render. Performance Analyzer shows that a single matrix accounts for almost all of it, and within that matrix the DAX query duration is about 15 seconds while Visual display and Other are each a few hundred milliseconds. Which conclusion does this evidence BEST support as the next thing to investigate?
- AThe semantic model behind the matrix: copy the recorded DAX query and examine how its measures, relationships, and cardinality evaluate.
- BThe on-premises data gateway, which is throttling the queries the matrix sends to the source and adding latency to every interaction on the page.
- CThe rendering cost of the matrix, most likely a slow custom visual or images being retrieved.
- DToo many visuals on the page, with the matrix waiting in a queue behind the other visuals before its own query can run.
Show the answer and explanation
Answer: A. The semantic model behind the matrix: copy the recorded DAX query and examine how its measures, relationships, and cardinality evaluate.
Correct. DAX query duration is the time between the visual sending its query and the semantic model returning results, so a figure that dominates the breakdown points at the measure expressions and model design.
Why not the others
- B. For imported tables a gateway is used only during refresh, not when visuals query the model; and where a table is in DirectQuery, the time spent at the external source, including the gateway hop, is reported separately as Direct query, not as DAX query.
- C. Drawing the visual on screen, including image retrieval and geocoding, is counted under Visual display. A few hundred milliseconds there rules this out.
- D. Time a visual spends waiting for other visuals or on background processing is reported under Other, which is negligible in this breakdown.
Takeaway: The value of Performance Analyzer is the category breakdown, not the total. DAX query points at the model and its measures, Direct query at the external source, Visual display at rendering, and Other at waiting and overhead.
Source: Use Performance Analyzer to examine report element performance
Product management wants to know which factors, such as customer role, subscription type, and device, lead customers to leave low ratings in a review table. Which visual is designed for this analysis?
- AA matrix of rating by role and device
- BA decomposition tree of ratings
- CA scatter chart of rating against device
- DA key influencers visual
Show the answer and explanation
Answer: D. A key influencers visual
Correct. The key influencers visual analyzes which factors drive a metric, including categorical outcomes such as low ratings.
Why not the others
- A. A matrix shows values but leaves the analyst to spot and rank influences by eye.
- B. A decomposition tree breaks a measure down level by level; it doesn’t rank which factors influence an outcome.
- C. A scatter chart compares two numeric values; device is categorical and only one factor fits.
Takeaway: Key influencers finds and ranks the factors associated with an outcome, for categorical or continuous metrics.
An analyst wants to put a customer revenue report on the company intranet so that only signed-in employees with access can see it, with RLS applied. A colleague suggests Publish to web. Why is that the wrong choice?
- APublish to web reports never refresh
- BPublish to web reports can’t be opened on phones
- CPublish to web requires a Premium capacity
- DAnyone on the internet can view it
Show the answer and explanation
Answer: D. Anyone on the internet can view it
Correct. With Publish to web, anyone on the Internet can view the published report, so confidential information must never be published that way.
Why not the others
- A. Refresh behavior is not what makes it unsuitable; exposure to the public is.
- B. Mobile access is not the concern here; the audience is.
- C. Capacity is not the issue; the report would be public.
Takeaway: Publish to web is public: anyone on the internet can view it. Use secure embedding options for internal or confidential content.
Source: Publish to web from Power BI
A Store dimension and a StoreTargets fact table both contain RegionCode and StoreNumber. Store numbers restart at 1 in each region, so StoreNumber alone is not unique, but the pair RegionCode and StoreNumber is. The analyst needs a one-to-many relationship from Store to StoreTargets. What should the analyst do?
- ARelate the tables on RegionCode and set cross-filter direction to Both.
- BIn both queries, add a column that combines RegionCode and StoreNumber, then relate on it.
- CCreate two relationships, one on RegionCode and one on StoreNumber, and keep both active.
- DRelate the tables on StoreNumber alone and set the cardinality to many-to-many.
Show the answer and explanation
Answer: B. In both queries, add a column that combines RegionCode and StoreNumber, then relate on it.
Correct. A Power BI relationship uses one column on each side, so a composite key is built as a single combined column in both tables, which is unique in Store.
Why not the others
- A. RegionCode is not unique in Store either, and cross-filter direction has no effect on which rows match. The store-level grain would be lost.
- C. Only one active relationship is allowed between two tables, and two separate single-column relationships would never behave like one relationship on the pair.
- D. The model would accept it, but store 1 in every region would match store 1 in every other region, so targets would be attributed to the wrong stores.
Takeaway: Model relationships join on a single column per table. When uniqueness depends on several columns, build one combined key column in both tables, ideally in Power Query.
FAQ
Which certification does PL-300 support?
It supports Microsoft Certified: Power BI Data Analyst Associate, as shown on Microsoft’s certification page.
How long is the PL-300 exam?
Microsoft’s certification page lists 100 minutes. Interactive components may appear, so read the exam experience page before test day.
Do I need to know DAX and Power Query?
Microsoft says candidates should be proficient with both. The outline covers measures, CALCULATE, time intelligence and many Power Query transformations.
What score do I need?
Microsoft states that a score of 700 or greater is required on its 1–1,000 scale, as described in its scoring policy.
How long should I study?
Use six weeks as a starting framework if you already work with data. Add time where your notes show gaps; no single schedule suits everyone.
Does the exam include Copilot topics?
The current outline lists several Copilot tasks in report creation and analysis. Review the published skills and give those items dedicated notes.
Next step
Try free practice questions → https://app.certmonk.app/samples
