Data Analyst Interview Questions: SQL, Excel and Behavioral Answers
Data analyst interview questions with sample answers: SQL, Excel, statistics, case and behavioral questions, plus push-back scripts and current BLS pay data.
On this pageWhat does a data analyst interview process look like?
Summary
Most data analyst interview questions fall into five groups: SQL, Excel or spreadsheets, statistics, business cases and behavioral stories. Expect a live SQL test on joins, GROUP BY and window functions, then a case like "sales dropped 10%, why?" Talk through your logic out loud, check your assumptions and tie every answer to a business decision. Clear communication often matters as much as the right query.
Data analyst interview questions test three things: whether you can pull the right data with SQL, whether you can find the story in it and whether you can explain that story to people who don't read code.
This guide covers the 16 questions you are most likely to hear, from joins and pivot tables to stakeholder conflict. Each comes with a sample answer or a clear plan.
You also get a case framework, push-back scripts and current pay data from the Bureau of Labor Statistics.
Quick answers by situation
| If they ask | What they want to hear | Example line |
|---|---|---|
| "What's the difference between a LEFT JOIN and an INNER JOIN?" | Clear logic plus a use case | "A LEFT JOIN keeps every customer, even those with no orders." |
| "When would you use a window function?" | Ranking or running totals per group | "I'd rank each region's top products without collapsing rows." |
| "How would you use XLOOKUP?" | Practical Excel skill | "I'd match order IDs to a price table, with exact match by default." |
| "Sales dropped 10%. Why?" | A structured plan | "First, I'd confirm the data is right, then split by region and channel." |
| "Tell me about a mistake in your analysis." | Ownership and a fix | "I caught a duplicate join, told my manager and added a row-count check." |
| "How do you explain results to non-technical people?" | One takeaway, one chart | "I lead with the decision, then show one chart that proves it." |
What does a data analyst interview process look like?
A typical data analyst interview process has four or five rounds: a recruiter screen, a hiring manager talk, a technical round on SQL and spreadsheets, a business case and a behavioral round. Some companies add a take-home project. SQL shows up in nearly every round, so practice it most.
A 2025 guide from Virginia Tech Career Services describes this same interview loop3. It also notes that an online SQL test may have 10 to 15 questions in 60 minutes.
SQL is non-negotiable in analytics interviews.
Virginia Tech Career and Professional Development, 2025
| Round | What they test | How to prepare |
|---|---|---|
| Recruiter screen | Fit, pay range, tools you know | A 60-second story and your salary range |
| Technical (SQL, Excel) | Joins, aggregation, cleaning | Timed practice problems |
| Business case | How you break down a vague problem | A repeatable framework |
| Take-home project | Analysis plus a short write-up | Clear charts and a one-page summary |
| Behavioral | Stakeholders, mistakes, influence | Five STAR stories |
Prepare for the screen first
The recruiter screen is often a quick check of tools and goals. So start with a crisp answer to "Tell me about yourself," and see our phone screen guide for the format.
What SQL questions come up in data analyst interviews?
The most common SQL questions in data analyst interviews cover joins, GROUP BY with HAVING, handling NULLs, removing duplicates and window functions like ROW_NUMBER and running totals. Interviewers want a correct query, but they also listen to how you explain it. So say your logic out loud before you type.
For broader tips on live coding rounds, see our guide to technical interview prep.
1. What is the difference between INNER JOIN and LEFT JOIN?
An INNER JOIN keeps only rows that match in both tables. A LEFT JOIN keeps every row from the left table, then fills missing matches with NULL.
2. What is the difference between WHERE and HAVING?
WHERE filters rows before grouping. HAVING filters groups after aggregation, for example customers with more than five orders.
3. How do you find duplicate rows?
Group by the columns that should be unique, then keep groups where COUNT(*) is greater than one. Next, explain how you would decide which copy to keep.
4. What is a window function?
PostgreSQL's documentation says a window function performs a calculation across a set of table rows5 related to the current row. Unlike GROUP BY, the rows keep their separate identities.
"I'd use ROW_NUMBER so each region keeps its own ranking."
SELECT region, product, revenue FROM (
SELECT region, product, revenue,
ROW_NUMBER() OVER (PARTITION BY region ORDER BY revenue DESC) AS rn
FROM sales) t
WHERE rn = 1;
5. How do you handle NULL values?
First, ask why they are missing. Then choose to filter them, replace them with COALESCE or flag them, and say how each choice changes the result.
Which Excel questions should you expect in a data analyst interview?
Excel questions in a data analyst interview usually cover pivot tables, lookups like XLOOKUP or INDEX-MATCH, data cleaning and simple charts. Some interviews hand you a messy sheet and a short time limit. Show a clean process: check the data, build the summary, then state one finding in plain words.
The Virginia Tech guide lists pivot tables and VLOOKUP versus INDEX-MATCH3 among common Excel topics. So expect at least one hands-on task. To describe the same skills on paper, see Rezoom's Excel skills examples for resumes.
6. How do you build a pivot table, and when would you use one?
Explain that a pivot table summarizes rows by category, such as revenue by month and region. Then give a real example from school or work.
7. VLOOKUP or XLOOKUP?
Microsoft's support page notes that VLOOKUP can only look from left to right4. XLOOKUP can search a column on either side, and it uses exact match by default.
"In my last [project or job], I used [XLOOKUP or a pivot table] to [combine two sheets or summarize sales].
That cut [the task] from [time before] to [time after].
Then I shared the result as [a chart or a one-page summary] with [the team]."
8. How would you clean a messy dataset?
Name your steps: remove duplicates, fix data types, trim spaces, standardize dates and check outliers. Also say you would keep the raw data untouched.
Narrate your checks
Before you share any number, say how you checked it, like "row counts match the source." Interviewers trust analysts who test their own work. It also gives you a natural way to catch your own mistakes out loud.
How do you answer statistics and case questions?
Statistics and case questions test how you think, not just what you know. Expect A/B testing, p-values, outliers and an open case like "sales dropped 10% last month." Use a simple structure: clarify the goal, check the data, segment the problem, test a few causes and recommend one next step.
The Virginia Tech guide lists hypothesis tests, sample size and Type I versus Type II errors3 as common statistics topics.
- Clarify the question Ask what changed, over what time frame and which metric matters most.
- Check the data Rule out a broken tracking tag, a missing day or a definition change.
- Segment the drop Split by region, channel, product and new versus returning customers.
- Recommend one action Pick the likely cause, name the next test and say what result would change your mind.
9. Sales dropped 10% last month. How would you find out why?
Walk through the steps above out loud. Also ask about seasonality, because last month may simply be slower every year.
10. Explain a p-value to a non-technical manager.
Keep it plain: a p-value shows how surprising the result would be if the change had no real effect. Then say what a small one means for the decision.
11. How do you handle outliers?
First, check whether the outlier is an error or a real event. Then decide to keep, cap or remove it, and explain how the choice moves the result.
Which behavioral data analyst interview questions should you expect?
Behavioral data analyst interview questions ask about stakeholder conflict, a mistake in your analysis, explaining results to non-technical people and working with messy data. Answer with the STAR method: situation, task, action and result. Always end with a business result, such as a decision made, time saved or revenue protected.
For the full method, read our guide to the STAR method. Class projects, internships and volunteer work all count for a first role.
12. Tell me about a time you found a mistake in your analysis.
"In a weekly sales report, I noticed revenue jumped 30% with no campaign behind it.
I traced it to a join that doubled some orders, fixed it and told my manager before the report went out.
Since then, I compare row counts before and after every join."
13. Tell me about a time a stakeholder disagreed with your findings.
Show that you listened, checked their concern with data and stayed open. Then share how you agreed on the next step.
14. How do you explain results to people who don't work with data?
Lead with the decision, show one chart and skip the jargon. Our guide to behavioral interview questions has more story prompts.
15. Tell me about a project you're proud of.
Pick one with a clear question, your own analysis and a real result. Even a class project works if a decision came out of it.
Related guideTechnical Interview Prep: A 2-Week Study Plan Built on Mocks, Not VolumeA technical interview prep plan for two weeks, day by day. Five mock interviews and mixed problem sets beat grinding 200 problems solo by topic.What to say if they push back
Interviewers often challenge your answer to see if you defend it with evidence or simply cave. Stay calm and lean on the data.
- Restate your logic. Walk through the steps that led to your answer.
- Name your assumption. Say what you assumed and how you would test it.
- Offer a check. Suggest one query or chart that would settle the question.
- Change your mind on evidence. Update your answer if they give you new facts, and say so.
[What if the drop is just seasonal?]
"That's possible, so I'd compare last month with the same month last year."
[And if it's down year over year too?]
"Then I'd segment by channel first, because a tracking change often hits one channel at a time."
How should you prepare for a data analyst interview?
Prepare for a data analyst interview by practicing timed SQL problems, rebuilding a pivot table and a lookup from scratch, rehearsing one case framework and writing five STAR stories. Study the company's product and guess its key metrics. Finally, prepare a two-minute walkthrough of one project with a clear business result.
Match your prep to the job posting. For example, the Virginia Tech guide says to prepare Python only if the job description calls for it3.
Tell one project story well
Pick one project and practice it in two minutes: the question, your data, your method and the result. If you're still shaping your resume, our resume objective for data analyst guide shows how to frame that same story on paper.
Check as you read
Before your data analyst interview
All 6 done. Next, check your resume against a real job post.
Check my resumeBack to this sectionSaved on this device
Guess the company's metrics
Before the interview, write down three metrics the company likely tracks, such as churn or order value. Then use one in your case answer. It shows business sense, which many analysts skip.
For questions to close with, see questions to ask in an interview.
What do data analyst pay and job numbers tell you?
The BLS has no profile titled "data analyst," so the closest matches are data scientists and operations research analysts. In May 2025, data scientists earned a median $120,230 a year, and operations research analysts earned $88,940. Data scientist jobs should grow 35% from 2025 to 2035.
Use these numbers to set your salary range. Also, entry-level analyst roles may pay less than these medians, so check local postings too.
| Role | Median pay (May 2025) | Growth 2025–35 | Openings per year |
|---|---|---|---|
| Data scientists | $120,2301 | 35% | About 24,800 |
| Operations research analysts | $88,9402 | 12% | About 7,500 |
Why growth shapes your answers
With 35% projected growth1, the analyst path leads to more technical roles. So mention one skill you're building, like Python or statistics, when they ask where you see yourself.
Do
- Talk through your SQL logic before you type.
- Check your numbers and say how you checked them.
- End every story with a business result.
- Ask clarifying questions in case rounds.
Don’t
- Jump to a query without restating the question.
- Hide a mistake in your analysis.
- Drown a manager in jargon or ten charts.
- Claim tools you can't use live.
After the interview, use our thank-you email guide to follow up. Then rehearse these data analyst interview questions out loud with Rezoom's AI mock interview.
Practice data analyst interview questions
with an AI interviewer
Frequently asked questions
What are the most common data analyst interview questions?
Common ones include the difference between joins, WHERE versus HAVING, how to find duplicates, how to build a pivot table and "sales dropped, why?" Behavioral questions about stakeholders and mistakes also come up often.
How much SQL do I need for a data analyst interview?
Know joins, GROUP BY with HAVING, subqueries, NULL handling and window functions like ROW_NUMBER. Practice under a timer, because many companies run a timed online test.
Do data analyst interviews include Python?
Some do. Prepare Python if the job description lists it; otherwise, focus on SQL, spreadsheets and statistics.
What is a data analyst take-home assignment?
It is a small project you complete on your own time, usually an analysis plus a short write-up or slides. Keep it focused on one clear recommendation.
How do I answer data analyst interview questions with no experience?
Use class projects, online courses, personal projects or volunteer work. Show the question, your method and the result, just as you would for a paid role.
What should I ask at the end of a data analyst interview?
Ask which tools the team uses, how analysts work with stakeholders and what a strong first 90 days looks like.
Sources
We opened and read each page below. The small numbers in the text point here. Pages that carry no publication date show n.d.
U.S. Bureau of Labor Statistics2026
Data scientists earned a median $120,230 a year in May 2025, with 35% projected growth from 2025 to 2035 and about 24,800 openings a year
bls.govU.S. Bureau of Labor Statistics2026
Operations research analysts earned a median $88,940 a year in May 2025, with 12% projected growth and about 7,500 openings a year
bls.govVirginia Tech Career and Professional Development2025
Data analyst interviews test SQL in nearly every round, and an online SQL test may have 10 to 15 questions in 60 minutes
career.vt.eduMicrosoft Supportn.d.
XLOOKUP uses exact match by default, while VLOOKUP can only look from left to right
support.microsoft.comPostgreSQL Global Development Group2026
A window function calculates across related rows while each row keeps its separate identity
postgresql.org