Interview questions
Power BI Interview Questions
20 Power BI interview questions with model answers: DAX, CALCULATE, star schema, measures, DirectQuery, RLS and report design. For both sides.

What's here
20 Power BI interview questions, grouped into 4 areas: data preparation and modelling, dax, connectivity, performance and security and reporting and stakeholder scenarios. Each one comes with why it is worth asking and what a strong answer contains, so the same question can be scored the same way by different interviewers.
- Give candidates a small dataset and ask them to sketch the star schema before building anything.
- Include one live DAX task, such as a year-over-year measure or share of total, and ask them to explain the filter context.
- Ask how they would secure a report so regional managers see only their region; check they know RLS does not apply to workspace editors.
Cohesyve finds, contacts and tests Power BI candidates on the real work, then ranks them for these interviews. 14-day free trial for 10 candidates, not charged until it ends.
Preparing for this interview yourself? Practise it free →These Power BI interview questions cover what data analyst and BI developer interviews test: shaping data with Power Query, star schema modelling and relationships, DAX measures and filter context, Import versus DirectQuery, row-level security, performance and report design, plus the stakeholder scenarios that decide whether a dashboard gets used. Candidates can use the 'what to look for' notes as model answers; interviewers can use them as a scoring guide.
Data Preparation and Modelling
Good Power BI work starts with the model. These questions test whether the candidate builds models that are fast and easy to use.
- 1
What is the difference between Power Query and DAX, and when do you use each?
Why ask it
Putting logic in the wrong layer causes slow, fragile reports.What to look for
Power Query (the M language) extracts, cleans and shapes data before it loads: merging, unpivoting, fixing types. DAX calculates on the loaded model: measures, calculated columns and tables. Rule of thumb: shape data as far upstream as possible and keep DAX for analysis that depends on report filters. - 2
What is a star schema, and why is it recommended for Power BI?
Why ask it
The single most important modelling concept.What to look for
Fact tables store events or observations with keys and numeric values; dimension tables describe entities such as products, customers and dates. Dimensions filter and group, facts are summarised. Microsoft's guidance recommends it for performance and usability. Strong candidates mention keeping fact tables at a consistent grain and avoiding one wide flat table. - 3
Explain relationship cardinality and filter direction. What happens when you need two relationships between the same tables?
Why ask it
Relationship mistakes produce wrong numbers silently.What to look for
One-to-many is the norm: the one side is the dimension, the many side the fact, with filters flowing from dimension to fact. Bidirectional filters should be used sparingly. Only one relationship between two tables can be active; others are inactive and used through USERELATIONSHIP in a measure, or by creating separate role-playing dimension tables such as Order Date and Ship Date. - 4
What is the difference between a calculated column and a measure?
Why ask it
A classic Power BI question that reveals depth quickly.What to look for
A calculated column is computed row by row and, in Import mode, stored in the model at refresh, which uses memory. A measure is calculated at query time in the current filter context and is never stored. Use measures for aggregations that respond to filters; use columns when you need a value to slice, filter or group by. - 5
Why do you need a dedicated date table?
Why ask it
Time intelligence depends on it.What to look for
A continuous date table with one row per day, marked as a date table, lets time intelligence functions work reliably and gives consistent year, quarter, month and fiscal attributes. Relying on auto date/time creates hidden tables and limits flexibility.
Before the Power BI interviews, see who can do the work
Cohesyve finds and contacts Power BI candidates from your job description and tests each one on an AI job simulation of the role, so these questions go to a ranked shortlist. 14-day free trial for 10 candidates, not charged until it ends.
DAX
Most Power BI interviews include DAX, and filter context is where candidates separate.
- 6
Explain row context and filter context.
Why ask it
The foundation of understanding DAX results.What to look for
Filter context is the set of filters applied to a calculation by slicers, visuals, rows and columns, and other measures. Row context exists when DAX evaluates row by row, as in calculated columns or iterators such as SUMX. Row context does not filter related tables on its own; that needs context transition. - 7
What does CALCULATE do?
Why ask it
The most important DAX function.What to look for
CALCULATE evaluates an expression in a modified filter context: filter arguments add new filters, or overwrite existing filters on the same columns, unless wrapped in KEEPFILTERS. Used without filters inside a row context, it performs context transition, turning the current row into a filter. Model measures referenced in row context transition automatically. - 8
Write a measure for sales in the same period last year and the year-over-year change.
Why ask it
A standard practical DAX test.What to look for
Sales LY = CALCULATE([Sales], SAMEPERIODLASTYEAR('Date'[Date])) and YoY % = DIVIDE([Sales] - [Sales LY], [Sales LY]). Strong candidates use an existing base measure, a proper date table, DIVIDE to handle zero or blank, and check behaviour for incomplete periods. - 9
What is the difference between SUM and SUMX?
Why ask it
Tests understanding of iterators.What to look for
SUM adds up one column. SUMX iterates a table, evaluates an expression for each row (such as quantity times price) and adds the results. Iterators are powerful but can be slow on large tables with complex expressions. - 10
How would you calculate each channel's share of total sales?
Why ask it
Tests control of filter context.What to look for
DIVIDE([Sales], CALCULATE([Sales], REMOVEFILTERS('Sales Order'[Channel]))), or ALL in older code. Strong answers explain which filters are removed and which remain, and consider whether the total should respect other slicers (ALLSELECTED).
Connectivity, Performance and Security
Choices that decide whether a report is fast, current and secure.
- 11
When would you use Import mode, DirectQuery or a composite model?
Why ask it
A core architecture decision.What to look for
Import loads data into Power BI's in-memory engine and is the default recommendation for speed and features, but data is only as fresh as the last refresh. DirectQuery leaves data in the source and queries it on every interaction, for very large or fast-changing data or source-enforced security, at the cost of performance and some modelling limits. Composite models mix the two, for example imported dimensions with DirectQuery facts. - 12
A report is slow. How do you find and fix the cause?
Why ask it
Performance troubleshooting is a daily task.What to look for
Use Performance Analyzer to find slow visuals and see whether time goes to DAX queries or rendering. Then check model size and column cardinality, remove unused columns, fix inefficient DAX, reduce visuals per page, and review relationships, especially bidirectional ones. In DirectQuery, check source queries and indexes. - 13
How does row-level security work in Power BI?
Why ask it
Data access mistakes are serious.What to look for
Roles are defined in Power BI Desktop with DAX filter expressions that return true or false per row, then users or groups are assigned in the service. Static rules use fixed values; dynamic rules compare USERPRINCIPALNAME() with a user-mapping table. RLS restricts users with Viewer permissions only: workspace Admins, Members and Contributors are not filtered. Test with 'Test as role'. - 14
What is query folding, and why does it matter?
Why ask it
Explains refresh performance and DirectQuery behaviour.What to look for
Query folding is when Power Query translates transformation steps into a native query that runs at the source, such as SQL. It reduces data transferred and speeds up refresh, and DirectQuery depends on it. Some steps break folding, so order steps carefully and check with View Native Query. - 15
How do you keep imported data up to date?
Why ask it
Stale dashboards erode trust.What to look for
Scheduled refresh in the service, an on-premises data gateway for on-premises sources, and incremental refresh for large fact tables so only recent partitions reload. Good answers mention monitoring refresh failures and telling users when data was last refreshed.
Reporting and Stakeholder Scenarios
A technically correct report that nobody trusts or uses has failed. These questions test the business side of the role.
- 16
How do you choose the right visual for a piece of information?
Why ask it
Design choices decide whether a report is understood.What to look for
Start from the question the user needs answered: line charts for trends over time, bar charts for comparisons, tables for exact values, cards for headline numbers. Avoid clutter, pie charts with many slices and decorative visuals; use consistent colours and clear titles that state the insight. - 17
How do you gather requirements for a new dashboard?
Why ask it
Many dashboards fail because they answer the wrong questions.What to look for
Ask who will use it, which decisions it supports and how often, which metrics matter and how they are defined, and where the data comes from. Agree metric definitions in writing, sketch a mock-up before building, and review early with real users. - 18
How do you share and manage Power BI content across a team?
Why ask it
Shows experience beyond Power BI Desktop.What to look for
Workspaces for development and collaboration, apps for distributing content to wider audiences, appropriate workspace roles, shared semantic models to avoid duplicated logic, and a promotion process from development to production, such as deployment pipelines where available. - 19
After publishing, a manager says a number on your dashboard is wrong. What do you do?
Why ask it
Tests ownership and communication.What to look for
Acknowledge quickly, find out which number and what they expected, reproduce it, trace it back through measures, model and source, and confirm the definition being used. Fix and communicate what changed, or explain the difference if the definitions differ, and add a check to stop it recurring. - 20
The sales total in Power BI does not match finance's Excel report. How do you reconcile them?
Why ask it
A very common real-world problem.What to look for
Compare definitions first (gross or net, returns, currency, date used, cut-off time), then filters and data freshness, then break both totals down by month, region or product to find where they diverge. Document the agreed definition so both reports use it.
Running the interview well
- 1Give candidates a small dataset and ask them to sketch the star schema before building anything.
- 2Include one live DAX task, such as a year-over-year measure or share of total, and ask them to explain the filter context.
- 3Ask how they would secure a report so regional managers see only their region; check they know RLS does not apply to workspace editors.
- 4Use the 'number is wrong' scenario to test communication and ownership as well as technical tracing.
- 5Look at a portfolio report if the candidate has one, and ask what they would change.
- 6Treat the PL-300 certification as a useful signal of structured study, but test modelling and DAX directly.
Common questions
What Power BI topics are asked most in interviews?
Power Query versus DAX, star schema, relationships, calculated columns versus measures, CALCULATE and filter context, time intelligence, Import versus DirectQuery, row-level security, performance and report design.
How should I prepare for a Power BI interview?
Build a report from a public dataset with a proper star schema, a date table, several DAX measures and row-level security. Practise explaining your model and measures out loud, and be ready to write DAX without autocomplete.
Is there a Power BI certification?
Yes. Microsoft offers the Microsoft Certified: Power BI Data Analyst Associate certification, earned by passing exam PL-300. It covers preparing, modelling, visualising and analysing data, and managing and securing Power BI. It is renewed every 12 months by passing a free online assessment on Microsoft Learn.
Do Power BI interviews include SQL?
Often, especially for analyst and BI developer roles, because data usually comes from SQL databases. Expect joins, aggregations and filtering.
What DAX functions should I know?
CALCULATE, FILTER, ALL and REMOVEFILTERS, SUMX and other iterators, DIVIDE, RELATED, USERELATIONSHIP, time intelligence functions such as SAMEPERIODLASTYEAR and TOTALYTD, and variables with VAR and RETURN.
Are Power BI interviews different for analysts and developers?
Analyst interviews weigh business questions, visual design and stakeholder communication. BI developer interviews go deeper into data modelling, performance, DirectQuery, deployment and security.
Save these questions for the Power BI candidates who can do the work
Cohesyve takes your Power BI job description, finds candidates, reaches out to them and runs an AI job simulation of the real work, then returns a ranked shortlist.
1,500+
assessments completed
50%
faster time-to-hire
5 min
from JD to assessment
14-day free trial · 10 candidates · Not charged until the trial ends
Preparing for a Power BI interview yourself? Practise on an AI job simulation of the role. 3-day free trial, cancel anytime.
From the blog