Power BI questions check how you turn data into reports people trust: data models, DAX measures and clear visuals.
Below are 30 Power BI questions with the correct answer and a short explanation, grouped from fresher to experienced level. Try each one before you open the answer.
In Tableau or Power BI, which of these fields would you normally treat as a measure rather than a dimension?
Answer: B. Revenue
Measures are numeric values you aggregate, such as summing revenue. Dimensions are the fields you slice by, such as region, category or date. Getting this right decides whether the tool adds values up or uses them to split the chart.
In Power BI you need to split a "City, Country" column, trim spaces and remove blank rows before modelling. Where should this happen?
Answer: B. In Power Query, during data load
Power Query (M) is the data preparation layer: shaping, cleaning and combining data as it loads. DAX is for calculations on the loaded model, such as measures and calculated columns. Cleaning in DAX is slower and harder to maintain.
A Power BI line chart shows months in the order April, August, December, February… How do you fix it?
Answer: C. Sort the month name column by month number
Text columns sort alphabetically by default. In the date table, select the month name column and use Sort by Column with a month number column (for an April–March fiscal year, a fiscal month number). Visuals then show months in the right order.
A sales manager wants to choose a region and see every chart on the page update. Which Power BI element is meant for this?
Answer: D. A slicer
A slicer is an on-canvas filter that users click to filter the other visuals on the page, and it can be synced across pages. Cards and KPI visuals display values; tooltips show details on hover.
A margin % measure is [Profit] / [Sales]. For new stores with zero sales the visual shows an error or infinity. What is the usual DAX fix?
Answer: B. Use DIVIDE([Profit], [Sales])
DIVIDE returns BLANK (or an alternate value you pass) when the denominator is zero, so visuals stay clean and blanks are skipped. It is also clearer than wrapping the division in IF checks.
A report published to the Power BI service uses a SQL Server inside the company network. Scheduled refresh fails. What is usually needed?
Answer: D. An on-premises data gateway
The cloud service cannot reach servers inside a private network on its own. An on-premises data gateway installed inside the network relays refresh and DirectQuery requests securely to those sources.
One Power BI report is shared with 20 regional managers, and each should only see data for their own region. Which feature is built for this?
Answer: C. Row-level security (RLS)
RLS defines roles with DAX filter rules, and the service applies them to every query for users in that role. One report and model then safely serves many audiences. Bookmarks and hidden columns do not restrict data.
A Power BI page takes 15 seconds to load. Which built-in tool shows how long each visual takes and its DAX query?
Answer: A. Performance analyzer
Performance analyzer records each visual's DAX query time, visual display time and other time when you refresh visuals. You can copy the query into DAX Studio to tune the slow measure.
On one Power BI page, users want a button to switch between a chart view and a table view of the same data. Which feature is usually used?
Answer: C. Bookmarks linked to buttons on the page
A bookmark saves the page state, including which visuals are visible. Two bookmarks, one showing the chart and one the table, linked to buttons, let users toggle views. Bookmarks also drive guided, story-like reports.
How does a product analytics tool such as Amplitude or Mixpanel differ from a general BI tool such as Power BI?
Answer: B. It is built for user events, funnels and cohorts
Product analytics tools model users and event sequences, so funnels, retention and paths need no SQL. BI tools are more general and great for business reporting across many sources, often built on a warehouse.
In Power BI you need Profit Margin = total profit / total revenue that responds to slicers for region and month. What should you create?
Answer: C. A measure such as DIVIDE(SUM(Sales[Profit]), SUM(Sales[Revenue]))
Measures are computed at query time in the current filter context, so the margin is correct for whatever slice the user picks. A calculated column is computed once per row at refresh, and averaging row-level margins gives the wrong total. DIVIDE also handles division by zero safely.
A Power BI report uses one huge flat table with 80 columns and is slow and hard to filter. What model is usually recommended?
Answer: C. A star schema: a fact table of transactions linked to dimension tables such as Date, Product and Customer
Power BI is designed around star schemas: narrow fact tables with numeric columns and keys, plus dimension tables for slicing. Filters flow cleanly from dimensions to facts, DAX is simpler and compression is better. A dedicated Date table also enables time intelligence functions.
A matrix has Region on rows. The measure below shows the same number on every row. Why?
UAE Sales = CALCULATE([Sales], Region[Region] = "UAE")
Answer: B. CALCULATE swaps the row's Region filter for UAE
A filter argument on a column in CALCULATE overwrites any existing filter on that column. On the India row, the Region filter is replaced by UAE, so every row returns UAE sales. Wrap it in KEEPFILTERS if it should intersect instead.
You want each product's share of total sales in a table with Product on rows, while still respecting a Year slicer. Which denominator is right?
Answer: A. CALCULATE([Sales], REMOVEFILTERS(Product))
REMOVEFILTERS(Product) (or ALL(Product)) removes only the product filter, so the denominator is total sales for the selected year. Removing every filter would also ignore the Year slicer. Then DIVIDE([Sales], denominator) gives the share.
A measure CALCULATE([Sales], SAMEPERIODLASTYEAR(Sales[OrderDate])) returns wrong or blank values for some months. What is the usual fix?
Answer: A. Use a continuous, marked date table instead
Time intelligence functions expect a date column with every day present. Fact table dates have gaps (no orders on some days), so shifting periods misbehaves. Use the date column of a continuous, marked date table in the function.
Sales has OrderDate and ShipDate, both related to the Date table, but only the OrderDate relationship is active. How do you build a "Sales by ship date" measure?
Answer: C. CALCULATE with USERELATIONSHIP on ShipDate
Only one relationship between two tables can be active. USERELATIONSHIP inside CALCULATE turns on the inactive one for that measure, so the Date table filters by ship date. The Date table is used as a role-playing dimension.
Users want fast, interactive Power BI reports on a 30 million row sales table, and data refreshed every morning is fine. Which storage mode suits best?
Answer: C. Import, refreshed on a schedule
Import loads the data into the compressed in-memory engine, which gives the fastest interaction, and a scheduled refresh keeps it current enough. DirectQuery is for data that must be near real time or is too big to import, at the cost of speed and source load.
A Power BI dataset holds 5 years of transactions and a full refresh takes 3 hours, though only the last few days change. What feature helps?
Answer: A. Incremental refresh with RangeStart/RangeEnd
Incremental refresh partitions the table by date using the RangeStart/RangeEnd filters in Power Query. Older partitions are kept and only recent ones are refreshed, cutting refresh time and source load. The filter should fold to the source.
A Power BI model is 2 GB, and most of it is a timestamp column with seconds. Reports only need date and hour. What reduces size the most?
Answer: B. Split it into a date and an hour column
The in-memory engine compresses columns by their distinct values. A timestamp to the second has millions of unique values; a date column has a few thousand and an hour column 24. Text would compress even worse.
Managers want a single chart where they choose to see Revenue, Orders or Margin %, instead of three separate charts. Which Power BI feature fits best?
Answer: A. Field parameters
Field parameters create a table of fields or measures that drives a slicer; the chosen one is swapped into the visual. They can switch dimensions too, such as region vs channel on the axis.
A Power BI table shows Distinct Customers by region: East 60, West 50. The total row shows 90, not 110. A manager thinks the report is broken. What is happening?
Answer: C. The measure is re-evaluated for all regions together, and 20 customers bought in both regions, so they are counted once
Measures are not summed for the total; they are recalculated in the total filter context. A distinct count over both regions counts each customer once, so overlapping customers make the total smaller than the sum. This is correct behaviour, but it should be explained or labelled for readers.
A table shows distinct customers: January 120, February 100. The total row shows 180, and a manager says it should be 220. Who is right?
Answer: D. The report; 40 customers bought in both months
A distinct count total counts each customer once across the whole period. If 40 customers bought in both months, the total is 120 + 100 − 40 = 180. Adding monthly distinct counts double-counts repeat customers.
This measure was meant to return all-time sales, but it returns the same as [Sales] for the current filter. Why?
All Time Sales =
VAR s = [Sales]
RETURN CALCULATE(s, ALL('Date'))Answer: B. s is computed once, before CALCULATE runs
A DAX variable is computed once in the filter context where it is defined, then treated as a constant. CALCULATE changes the context for expressions it evaluates, but s is already a number. Write CALCULATE([Sales], ALL('Date')) instead.
A matrix has Product Color on rows. Measure A = CALCULATE([Sales], Product[Color] = "Red") shows Red sales on every row. What does KEEPFILTERS(Product[Color] = "Red") change?
Answer: B. Only the Red row shows a value; others are blank
Without KEEPFILTERS, the new filter replaces the row's colour filter, so every row becomes "Red". KEEPFILTERS intersects the filters instead: on the Blue row, Blue AND Red is empty, so it is blank, and only the Red row returns sales.
In a card filtered to 2026, which measure returns the average sales per month, not per transaction?
Answer: C. AVERAGEX(VALUES('Date'[Month]), [Sales])
AVERAGEX over the distinct months computes [Sales] for each month (through context transition) and averages those monthly totals. AVERAGE(Sales[Amount]) and the other two give the average transaction value, which is a much smaller number.
In the Customer table, a calculated column Total = SUM(Sales[Amount]) shows the grand total of all sales on every row. Changing it to CALCULATE(SUM(Sales[Amount])) gives each customer's total. Why?
Answer: A. CALCULATE turns the row into a filter
A calculated column has a row context but no filter context, so SUM sees all of Sales. CALCULATE performs context transition: the current customer row becomes a filter, which flows through the relationship to Sales. Measures called inside a row context do this automatically.
A Power BI table visual in DirectQuery mode lists every transaction line and fails with an error about too many rows. What limit is being hit?
Answer: B. The 1 million row DirectQuery result cap
In DirectQuery, Power BI limits how many rows an intermediate query may return (1 million by default). Huge detail tables hit it. Aggregate in the visual, filter more, or use drill-through for detail.
A matrix of Store × Product used to show only products each store sold. After changing a measure to [Sales] + 0, it lists every product for every store and gets slow. Why?
Answer: D. Blank rows are hidden, but 0 is not blank
Power BI drops row combinations where all measures return BLANK. Forcing 0 makes every store-product pair non-blank, so the visual shows the full cross join and computes far more cells. Return BLANK for no data unless zeros are truly needed.
RLS works in testing, but one regional manager sees all regions in the Power BI service. She is a Member of the report's workspace. Why?
Answer: A. RLS does not apply to workspace Members
Workspace Admins, Members and Contributors can edit the content, so RLS does not restrict them. Give report consumers the Viewer role, or share through an app, so RLS rules apply.
In a pivot with Region on rows, what does this DAX measure return for each region?
Answer: A. The region's sales as a share of all regions' sales
The numerator uses the current filter (one region). CALCULATE with ALL(Sales[Region]) removes the region filter, so the denominator is the total for all regions. DIVIDE also returns blank instead of an error when the denominator is zero.
Take a 15-question Power BI MCQ test with a score at the end, or practise the full Power BI question bank and earn points on the leaderboard.