12 DAX Interview Questions (With Answers That Impress)

Where Interviews Are Won and Lost

Most candidates can handle "what is Power BI?". The questions that filter people out are DAX questions — because DAX tests whether you truly understand how Power BI thinks, not just whether you've clicked around. Here are the 12 DAX questions that appear again and again, with the answers that separate strong candidates from average ones.

Q1. What is the difference between a measure and a calculated column?

A measure is evaluated at query time in filter context and isn't stored — the right tool for business metrics. A calculated column is evaluated at load time, stored in memory, and adds a physical column. Measures are preferred for most calculations because they're dynamic, memory-efficient and respect context.

Q2. Explain filter context vs row context.

Filter context is the environment of filters (slicers, page filters, CALCULATE) that determines which rows are visible for a calculation. Row context is the current row during iteration — in calculated columns, SUMX, FILTER and other iterators. The link between them is context transition, performed by CALCULATE when it's used inside row context.

Q3. When would you use CALCULATE vs FILTER?

CALCULATE with simple filter expressions is preferred for performance — the engine handles these efficiently. FILTER is an iterator needed only when the condition can't be a simple column filter, such as filtering by a measure value or a complex computation. Simple first, FILTER as the exception.

Q4. What is context transition?

Context transition is the conversion of row context into filter context that happens when CALCULATE (or a measure reference, which implies CALCULATE) is evaluated inside row context. It lets a row-level computation become a filterable calculation — the mechanism behind many subtle DAX behaviours.

Q5. How do you calculate YoY growth?

Build a proper date table first. Then: Sales LY = CALCULATE([Total Sales], SAMEPERIODLASTYEAR('Date'[Date])) and YoY Growth = DIVIDE([Total Sales] – [Sales LY], [Sales LY]). Mention TOTALYTD for year-to-date comparisons and the need to mark the date table.

Q6. What does DIVIDE do and why use it?

DIVIDE(a, b, [alternate]) safely divides by zero, returning BLANK (or the alternate) instead of an error. The bare "/" operator can produce DIV/0 or infinity that silently corrupts downstream calculations. Professional models use DIVIDE everywhere.

Q7. What is ALL and how does it differ from ALLEXCEPT?

ALL removes all filters from a table or column — used for "share of total" ratios. ALLEXCEPT removes all filters except those on specified columns, useful when you want to keep certain filters while removing others. Both are typically used inside CALCULATE to redefine context.

Q8. How do you rank products in context?

Use RANKX: Rank = RANKX(ALL(Product), [Total Sales], , DESC). Inside ALL(Product) the ranking ignores product-level filters so the ranking is stable across the whole list, while [Total Sales] evaluates in the current context.

Q9. What are iterator functions and when do you need them?

Iterators (SUMX, AVERAGEX, MAXX, MINX, RANKX) evaluate an expression row by row and aggregate the results. You need them when a calculation depends on a per-row value that isn't a stored column — like quantity × price, or a row-level margin.

Q10. What is the difference between SUMX and SUM?

SUM adds a single existing column. SUMX iterates a table, evaluates an expression per row, and sums those results. Use SUM when the column already holds the value; use SUMX when the value must be computed per row.

Q11. How do you debug a DAX measure that returns the wrong total?

Break it down: check the source data and types, verify relationships and cross-filter direction, test the measure in a matrix with multiple dimensions to expose context problems, and isolate each component (CALCULATE filter, time intelligence, DIVIDE) separately. Many "wrong totals" turn out to be filter-context misunderstandings, not math errors.

Q12. How would you optimize a DAX-heavy report?

Model first: star schema, reduce cardinality, remove unused columns. Then DAX: prefer simple CALCULATE filters over FILTER, store repeated subexpressions in variables (VAR…RETURN), avoid context transition in huge tables, and check DAX Studio for the worst queries. Then operational: incremental refresh and aggregation tables.

DAX interviews reward understanding, not memorization. If you can explain why, you'll pass.

The Three Answers That Win the Room

Across all twelve questions, interviewers are really testing three themes. First, context — can you explain how filters flow through a model and what changes when a user clicks? Second, iteration — do you understand when the engine walks row by row and why that changes performance and results? Third, purpose — can you connect every function back to a business problem, rather than reciting syntax? When you answer with all three in mind, even unfamiliar questions become manageable, because you're reasoning from principles instead of recalling lines.

That's the hidden pattern behind the "impressive" answers interviewers remember: they're not memorized scripts, they're structured reasoning applied on the spot. A candidate who says "I'd use CALCULATE here because it modifies filter context without iterating" sounds senior because they're demonstrating understanding, not recall. Rehearse the twelve questions, but rehearse the reasoning out loud — the why matters more than the what.

How to Prepare in Seven Days

• Days 1–2: rebuild every answer from memory, then check against these models.
• Days 3–4: build one small model and physically demonstrate CALCULATE, context transition and DIVIDE.
• Day 5: record yourself answering five questions and listen for vagueness.
• Day 6: run a mock interview with a friend or mentor under real time pressure.
• Day 7: review your notes, then rest — a fresh mind answers better than a crammed one.

The best preparation is a mix of understanding and rehearsal, which is exactly how our mock-interview sessions work: hard questions, real pressure, honest feedback, and the chance to hear how your answers sound to an experienced interviewer. After one or two such sessions, the twelve questions here won't feel like a list to memorize — they'll feel like familiar ground you've already walked.

And when the interviewer asks the final question — "do you have anything to ask us?" — use it: ask what their data models look like, how the team handles governance, or what a successful first quarter looks like. Informed questions are the quiet DAX of the interview: they show understanding without a word of syntax, and they're how strong candidates end strong.

Walk in having practised the twelve, understood the three themes, and prepared your own questions — and you'll find the DAX round is no longer the one you fear, but the one you're ready for.

Our Top Power BI Course in India

Looking for Power BI training in a specific city? Explore our city-wise Power BI Course pages to find course information, training details, curriculum, practical learning approach, and career guidance for your location.