50 Power BI Interview Questions and Answers (2026), With DAX Examples

Last updated: October 2026

Power BI interviews look easy from the outside. "It's just dashboards, right?" Then the interviewer asks why your measure shows the same number in every row of the table, and the room goes quiet.

That's the real test. Anyone can drag a bar chart onto a canvas. Companies hire people who understand the data model underneath it: relationships, filter context, and how DAX actually calculates. This guide covers 50 questions that come up for Power BI developer, data analyst and BI analyst roles in India, from basics to the scenario questions senior candidates get.

Power BI interview questions: the quick answer

• What gets tested: star schema modelling, relationships and cardinality, Power Query (M), DAX (CALCULATE, filter context, time intelligence), row-level security, publishing and refresh, performance.
• Typical format: a theory round, a DAX round (write 3–5 measures live), and often a take-home or live task: "here's a sales file, build me a one-page report in 45 minutes".
• The concept that decides most interviews: row context vs filter context, and how CALCULATE changes filter context.
• Bring proof: a public Power BI report or a short video walkthrough of a dashboard you built beats any certificate.

How Power BI interviews actually run

Service companies and captive centres (GCCs) usually start with tool questions: what's Power Query, what's a gateway, how do you schedule refresh. Analytics consultancies and product companies jump to DAX and modelling fast, and many ask you to share your screen and build something.

We use one simple model in the DAX examples: a Sales fact table (OrderDate, ProductKey, CustomerKey, Quantity, Amount) linked to Date, Product and Customer dimension tables.

Basic Power BI interview questions (1–15)

1. What is Power BI?

Microsoft's business intelligence tool for connecting to data, cleaning it, modelling it and building interactive reports and dashboards. It has three main parts: Power BI Desktop (build), the Power BI Service (share and refresh), and Power BI Mobile (view). In 2026 it also sits inside Microsoft Fabric, Microsoft's wider analytics platform.

2. What are the main components of Power BI?

• Power Query: gets and transforms data.
• Data model (VertiPaq engine): stores data in compressed columns and holds relationships.
• DAX: the formula language for measures and calculated columns.
• Report view: visuals and pages.
• Power BI Service: workspaces, apps, dashboards, scheduled refresh and sharing.

3. What is the difference between a report and a dashboard?

A report is one or more pages of interactive visuals built on one semantic model (dataset). A dashboard is a single page in the Power BI Service made of tiles pinned from one or more reports. Dashboards are for at-a-glance monitoring; reports are for exploring.

4. What data sources can Power BI connect to?

Hundreds: Excel, CSV, SQL Server, PostgreSQL, MySQL, Oracle, SharePoint, Azure SQL, Snowflake, BigQuery, Databricks, Salesforce, web APIs, and Fabric lakehouses and warehouses, among others.

5. What are Import, DirectQuery and Live Connection modes?

There's also Direct Lake in Fabric, which reads Delta tables in OneLake without importing them. Mention it if the role uses Fabric.

6. What is Power Query?

The data preparation layer. You use it to remove columns, fix data types, split columns, unpivot, merge and append tables. Every step is recorded and written in the M language, so refreshes repeat the same steps automatically.

7. What is the difference between Merge and Append in Power Query?

Merge joins two tables side by side on a key, like a SQL join. Append stacks tables on top of each other, like SQL UNION ALL. For example, append January and February sales files; merge Sales with Product to bring in category.

8. What is DAX?

Data Analysis Expressions, the formula language Power BI uses for measures, calculated columns and calculated tables. It looks a bit like Excel formulas but works on whole tables and columns, and it's sensitive to filter context.

9. What is the difference between a measure and a calculated column?

Rule of thumb: if you'll put it on an axis or slicer, it's probably a column. If you'll show it as a number, make it a measure.

10. What is a star schema?

A model with one central fact table (numbers and keys, such as sales transactions) surrounded by dimension tables (descriptions, such as date, product, customer, region). Power BI is designed for star schemas: they're smaller, faster and make DAX simpler.

11. What is the difference between fact and dimension tables?

Fact tables record events and measurements and usually have many rows (orders, clicks, payments). Dimension tables describe the who, what, where and when, and usually have fewer rows (customers, products, dates).

12. What types of relationships does Power BI support?

One-to-many (the most common), one-to-one and many-to-many. Each relationship also has a cross-filter direction: single (the dimension filters the fact) or both. Default to one-to-many with single direction.

13. What are slicers?

On-canvas filters users click to filter a report page, for example a dropdown for region or a date range slider. They can be synced across pages.

14. What are the different types of filters in Power BI?

Visual-level, page-level and report-level filters in the filter pane, plus slicers, cross-filtering between visuals, drillthrough filters and row-level security (which filters data by user).

15. What is a Power BI gateway?

Software that securely connects the Power BI Service in the cloud to data sources sitting on-premises, like a SQL Server in your office, so that scheduled refresh and DirectQuery can work. The standard mode is shared by many users; personal mode is for one person.

Intermediate Power BI and DAX questions (16–35)

16. What is filter context?

The set of filters applied to a calculation at a given moment: slicers, rows and columns of the visual, page and report filters. A [Total Sales] measure shows a different number in each row of a table because each row has a different filter context.

17. What is row context?

The "current row" when DAX goes through a table row by row, which happens in calculated columns and in iterator functions like SUMX. Row context does not filter other tables on its own. That's what context transition (via CALCULATE) is for.

18. What does CALCULATE do?

It evaluates an expression in a modified filter context. It's the most important function in DAX.

Sales North = CALCULATE ( [Total Sales], Region[Region] = "North" )

CALCULATE also triggers context transition: when it's used inside a row context, it turns the current row into an equivalent filter.

19. What is the difference between SUM and SUMX?

SUM adds up one column. SUMX iterates a table, evaluates an expression for each row, then adds the results.

Total Sales = SUM ( Sales[Amount] )
Revenue = SUMX ( Sales, Sales[Quantity] * RELATED ( Product[Price] ) )

Use SUMX when the value has to be calculated per row first.

20. What is the difference between ALL, ALLEXCEPT and ALLSELECTED?

• ALL(table or column) removes all filters from it.
• ALLEXCEPT(table, column) removes all filters from the table except those on the listed columns.
• ALLSELECTED() removes filters coming from the visual but keeps filters from slicers and the page. It's useful for "% of what the user has selected".

21. Write a measure for percentage of total sales.

Sales % of Total =
DIVIDE (
    [Total Sales],
    CALCULATE ( [Total Sales], ALL ( Product[Category] ) )
)

Use DIVIDE instead of /. It handles division by zero safely.

22. What is the difference between FILTER and CALCULATE filters?

A simple boolean filter inside CALCULATE (Product[Color] = "Red") is efficient. FILTER() returns a table by checking a condition row by row, so it's needed for complex conditions, such as filtering on a measure, but it's slower on big tables. Filter columns, not whole fact tables, when you can.

23. What are time intelligence functions?

DAX functions for date comparisons: TOTALYTD, DATESYTD, SAMEPERIODLASTYEAR, DATEADD, PARALLELPERIOD, DATESINPERIOD. They need a proper date table marked as a date table, with one row per day and no gaps.

24. Write year-to-date and year-over-year growth measures.

Sales YTD = TOTALYTD ( [Total Sales], 'Date'[Date] )

Sales LY = CALCULATE ( [Total Sales], SAMEPERIODLASTYEAR ( 'Date'[Date] ) )

YoY Growth % = DIVIDE ( [Total Sales] - [Sales LY], [Sales LY] )

For an Indian financial year ending 31 March, pass the year-end date: TOTALYTD([Total Sales], 'Date'[Date], "31/3").

25. Why do you need a separate date table?

Time intelligence functions need a continuous range of dates. A date table also lets you add columns the business uses, such as financial year, quarter, month name and week number, and keeps one consistent calendar across all fact tables. Turn off Auto date/time in big models; it creates hidden tables for every date column.

26. What is RELATED vs RELATEDTABLE?

RELATED fetches a value from the "one" side of a relationship into the "many" side (for example, Product[Category] into a Sales calculated column). RELATEDTABLE goes the other way and returns the matching rows from the "many" side.

27. What are variables in DAX, and why use them?

VAR stores an intermediate result, and RETURN gives the final value. Variables make measures readable, avoid calculating the same thing twice, and help with debugging.

Margin % =
VAR Revenue = [Total Sales]
VAR Cost = SUM ( Sales[Cost] )
RETURN DIVIDE ( Revenue - Cost, Revenue )

28. What is the difference between COUNT, COUNTA, COUNTROWS and DISTINCTCOUNT?

COUNT counts non-blank numbers (and dates) in a column, COUNTA counts non-blank values of any type, COUNTROWS counts rows in a table, and DISTINCTCOUNT counts unique values. Use COUNTROWS(Sales) for order counts; it's clearer and fast.

29. How do you rank products by sales in DAX?

Product Rank =
RANKX ( ALL ( Product[ProductName] ), [Total Sales], , DESC, DENSE )

ALL is the key part. Without it, each row only sees itself and every product gets rank 1. That's a favourite trick question.

30. What is a role-playing dimension, and how do you handle it?

One dimension used in several roles, for example a Date table that needs to relate to both OrderDate and ShipDate. Power BI allows only one active relationship between two tables, so you keep the second one inactive and switch to it inside a measure:

Sales by Ship Date =
CALCULATE ( [Total Sales], USERELATIONSHIP ( Sales[ShipDate], 'Date'[Date] ) )

31. What is row-level security (RLS)?

A way to restrict which rows each user can see. You create roles with DAX filters in Desktop, for example [Region] = "South", then assign users or groups to roles in the Service. Dynamic RLS uses USERPRINCIPALNAME() to match the logged-in user against a mapping table, so you don't need one role per region.

32. What is the difference between static and dynamic RLS?

Static RLS hard-codes a filter per role (one role per region). Dynamic RLS uses a single role whose filter compares a user-email column to USERPRINCIPALNAME(), so access is managed in a table instead of in Power BI roles. Dynamic RLS scales much better.

33. What are bookmarks?

Saved states of a report page: filters, slicers, visible visuals and drill state. They're used for storytelling, "reset filters" buttons, and toggling between chart and table views.

34. What is drillthrough?

A feature where a user right-clicks a data point, such as a customer, and jumps to a detail page already filtered to that customer. You set it up by adding a field to the drillthrough well of the target page.

35. How does scheduled refresh work?

In the Power BI Service, you set credentials for the semantic model, connect a gateway if the data is on-premises, and choose refresh times. Pro licences allow up to 8 scheduled refreshes a day; Premium and Fabric capacities allow up to 48. Incremental refresh lets large tables refresh only recent partitions.

Advanced and scenario-based Power BI questions (36–50)

36. How do you improve the performance of a slow Power BI report?

Work from the model outward:

1. Use a star schema; remove unused columns, especially high-cardinality text and IDs.
2. Prefer Import over DirectQuery where freshness allows.
3. Replace calculated columns with Power Query steps or source-side logic.
4. Optimise heavy DAX: use variables and avoid FILTER on whole fact tables.
5. Reduce visuals per page; each one sends its own query.
6. Use Performance Analyzer in Desktop and DAX Studio to find the slow visual or measure.

37. What is Performance Analyzer?

A pane in Power BI Desktop that records how long each visual takes to render, split into DAX query time, visual display time and other. You can copy the DAX query into DAX Studio to dig deeper.

38. What is incremental refresh?

A policy that partitions a large table by date, so each refresh reloads only recent data (say, the last 7 days) and keeps older partitions as they are. It needs RangeStart and RangeEnd parameters in Power Query and works best when the source supports query folding.

39. What is query folding?

When Power Query steps are translated into a single native query, such as SQL, and run at the source instead of inside Power BI. Folding makes refresh much faster. Some steps break folding (for example, adding an index column early), so do foldable steps first. Right-click a step and check "View Native Query" to confirm.

40. What is a composite model?

A model that mixes storage modes, for example a large fact table in DirectQuery with small dimensions in Import, or a local model that extends a published semantic model. Aggregation tables in Import mode can then answer most queries fast, while detail queries go to DirectQuery.

41. What are calculation groups?

A feature that lets you define a set of calculation items, like YTD, MTD, Last Year and YoY %, once, and apply them to any measure. Without them you'd write 4 time-intelligence variants of every measure. You can create them in Desktop's model view or in Tabular Editor.

42. What is the difference between a semantic model, a dataflow and a datamart?

A semantic model (formerly "dataset") is the model reports run on. A dataflow is Power Query in the cloud that prepares reusable tables for many models. A datamart combined a managed database with a model for self-service users; Microsoft has been steering new work towards Fabric warehouses and lakehouses instead.

43. Why should you avoid bi-directional relationships?

They can create ambiguous filter paths, give unexpected numbers and slow queries down. Use single direction by default and turn on both-way filtering only where needed, or apply it inside a specific measure with CROSSFILTER().

44. Your measure shows the same total in every row of a table. Why?

Usually because the field on the rows has no relationship to the table the measure aggregates, or the relationship is inactive or filters in the wrong direction. Check the model view first. Less often, the measure uses ALL() and removes the filter on purpose.

45. A manager wants sales for the last 3 months to always show, whatever the date slicer says. How?

Sales Last 3M =
VAR MaxDate = MAX ( 'Date'[Date] )
RETURN
    CALCULATE (
        [Total Sales],
        DATESINPERIOD ( 'Date'[Date], MaxDate, -3, MONTH )
    )

Follow up with a question: should "last 3 months" end at today or at the last date with data? Clarifying requirements is part of the answer.

46. How would you show "Top 5 products, with everything else as Others"?

Create a disconnected table of product names plus an "Others" row, rank products with RANKX, and write a measure that returns sales for the top 5 by name and the sum of the rest for "Others". Mention that the built-in Top N filter on a visual shows the top 5 but can't add an "Others" bar.

47. How do you handle a many-to-many relationship, such as customers with multiple accounts?

Ideally, add a bridge table (CustomerAccount) between the two dimensions, so both relationships are one-to-many. Power BI's native many-to-many cardinality works too, but bridge tables make the logic easier to reason about and test.

48. How do you manage dev, test and production versions of reports?

Use separate workspaces for each stage and promote content with deployment pipelines. In Fabric, Git integration lets you keep report and model definitions (PBIP format) in Azure DevOps or GitHub. Parameters switch data sources between environments.

49. How do you make sure the numbers in your dashboard are right?

Reconcile totals with the source system or finance's official figure, test edge cases (no sales, returns, a new product with no history), compare a few rows by hand, and document measure definitions so "revenue" means one thing everywhere. Interviewers love this question because many candidates never validate.

50. Walk me through a dashboard you built.

This is the question that decides senior interviews. Structure your answer:

1. Business question: who used it and what decision it supported.
2. Data: sources, volume and refresh frequency.
3. Model: fact and dimension tables and any tricky relationships.
4. Key measures: 2–3 DAX measures you're proud of.
5. Impact: what changed, such as hours saved per week or a decision made faster.

Keep it to about two minutes, then let them ask questions.

Common mistakes candidates make

• Building one giant flat table instead of a star schema.
• Using calculated columns for things that should be measures.
• Not knowing why RANKX without ALL returns 1 everywhere.
• Saying "I used CALCULATE" without being able to explain filter context.
• Showing a dashboard with 15 visuals on one page. Simple and clear wins.

A 7-day Power BI prep plan

Microsoft's free Learn path for the PL-300 (Power BI Data Analyst) exam covers most interview topics. The certificate helps with shortlisting at service companies, but a real project gets you through the interview.

Which roles ask Power BI, and what they pay

Power BI is core for BI developers, data analysts, MIS analysts, business analysts and finance analysts. For pay by experience and city, see our data analyst salary guide and business analyst salary guide. Company-wise numbers are on Weekday's salary pages.

Nearly every Power BI interview also includes SQL, so pair this with our SQL interview questions. Also useful: BI developer interview questions and Power BI developer resume examples.

Frequently asked questions

Is Power BI hard to learn?

The interface is easy; most people build a basic report in a day. DAX and data modelling take longer, usually a few weeks of practice, because filter context works differently from Excel.

Is DAX asked in fresher Power BI interviews?

Yes. Expect at least basic measures (SUM, CALCULATE, DIVIDE), the measure vs calculated column difference, and one time-intelligence measure like YTD.

Should I learn Power BI or Tableau?

In India, Power BI has more job openings, especially at service companies, GCCs and companies already on Microsoft 365. Tableau is still common in some product and consulting firms. Concepts carry over, so start with Power BI.

Is the PL-300 certification worth it?

It helps your resume get shortlisted, especially at IT services companies and Microsoft partners. It won't replace a portfolio. Do both if you can.

What is Microsoft Fabric, and do I need it for Power BI interviews?

Fabric is Microsoft's analytics platform that brings Power BI, data engineering, warehousing and real-time analytics together on OneLake. For pure reporting roles, basic awareness is enough. For BI engineer roles, expect questions on lakehouses, Direct Lake and dataflows.

Ready to apply? Browse data and BI jobs on Weekday and check your resume with our free ATS resume checker first.