Getting the interview is the hard part — but plenty of capable candidates still walk out of one having under-sold themselves. Not because they lacked the skills, but because they weren't ready for the shape of the questions. Data analyst interviews follow a fairly predictable pattern, and once you know it, you can prepare for it.
This guide covers 30 of the most common data analyst interview questions, grouped into the four types you'll actually face: technical and SQL, analytical and case, behavioural, and the project walkthrough. For the key ones, you'll get a sample answer or a clear approach — so you're rehearsing real responses, not just reading a list.
1. Technical & SQL Questions
These test whether you can actually do the day-to-day work. For junior roles, SQL is almost always the focus.
Q1. What's the difference between WHERE and HAVING in SQL?
Approach: WHERE filters rows before aggregation; HAVING filters groups after aggregation (used with GROUP BY). A clean example beats a definition: "I'd use WHERE to filter to 2025 orders, then GROUP BY customer, then HAVING to keep only customers with more than 10 orders."
Q2. Explain the different types of JOIN.
Approach: Cover INNER (matching rows in both tables), LEFT (all rows from the left table plus matches), RIGHT (the reverse), and FULL OUTER (everything from both). Mention that LEFT JOIN is the one you reach for most when checking what's missing from the second table.
Q3. How would you find duplicate records in a table?
Approach: GROUP BY the columns that define a duplicate, then use HAVING COUNT(*) > 1. Show you understand what "duplicate" even means in context — exact row copies vs. the same customer entered twice.
Q4. Write a query to find the second-highest value in a column.
Approach: Several valid routes — a subquery with MAX, or a window function like DENSE_RANK(). Say your approach out loud as you write. Interviewers care more about your reasoning and how you handle ties than a perfectly memorised one-liner.
Q5. What's the difference between DELETE, TRUNCATE and DROP?
Approach: DELETE removes rows (and can be filtered with WHERE); TRUNCATE removes all rows fast but keeps the table; DROP removes the table entirely. Knowing this signals you understand you're working with real, consequential data.
Also be ready for:
- What is a primary key vs a foreign key?
- What does GROUP BY do, and when do you use aggregate functions?
- How would you handle NULL values in a query?
- What Excel functions do you use most? (VLOOKUP/XLOOKUP, INDEX-MATCH, pivot tables, SUMIFS)
- Which visualisation tool do you know, and how do you decide on a chart type?
2. Analytical & Case Questions
These test how you think. There's rarely one correct answer — they want to see structured reasoning.
Q6. Our sales dropped 20% last month. How would you investigate why?
Approach: Don't guess a cause — show a method. Segment the drop (which product, region, channel, customer type?), check whether it's a data issue first, compare against seasonality and last year, then isolate the biggest contributor. Structured curiosity is exactly what they're testing.
Q7. How would you measure the success of a new feature?
Approach: Start by clarifying the goal, then define a primary metric tied to it (e.g. adoption or retention), plus guardrail metrics to catch side-effects. Naming a metric and why it's the right one shows business thinking, not just number-crunching.
Q8. How do you check whether your analysis is correct?
Approach: Sanity-check totals against a known source, look for impossible values, test edge cases, and get a second pair of eyes. Showing that you don't blindly trust your own output is a strong signal of maturity.
Q9. How would you explain a complex finding to a non-technical stakeholder?
Approach: Lead with the answer and the "so what", drop the jargon, and use one clear visual. This is where you can mention leading with the conclusion — a principle covered in our post on the Pyramid Principle for data analysts.
Also be ready for:
- How would you decide which metric matters most for a business?
- What would you do if two data sources disagreed?
- How do you prioritise when you have several analysis requests at once?
- What's the difference between correlation and causation?
3. Behavioural Questions (Use the STAR Method)
These explore how you've worked in the past. The best way to answer them is the STAR method — Situation, Task, Action, Result.
STAR keeps your answers focused and evidence-based instead of vague. You set the scene, explain what you were responsible for, describe what you did, and state the outcome. It's the standard framework recommended by the UK's National Careers Service STAR method guide, and it works just as well when you're drawing on projects, university, or a previous unrelated job.
Q10. Tell me about a time you worked with a difficult dataset.
Approach: Pick a real project. Situation: the messy data. Task: what you needed from it. Action: how you cleaned and handled it. Result: what you delivered and learned.
Q11. Describe a time you made a mistake in your analysis.
Approach: Don't claim you've never made one — that reads as either dishonest or inexperienced. Show a real, small mistake, how you caught it, and what process you changed so it wouldn't happen again. Self-awareness scores well here.
Q12. Tell me about a time you had to explain data to someone who disagreed with it.
Approach: Focus on how you stayed calm, listened, and used the data to find common ground rather than "winning". Communication under mild conflict is a genuinely valued analyst skill.
Also be ready for:
- Tell me about a time you managed a tight deadline.
- Describe a project you're proud of and why.
- Give an example of when you had to learn a new tool quickly.
- Tell me about a time you worked in a team.
- Why do you want to be a data analyst?
4. The Project Walkthrough
Almost every data interview includes "walk me through a project." For a candidate with no job experience, this is the most important question of all — your project is your experience.
Q13. Walk me through a project from your portfolio.
Approach: Structure it: the question you set out to answer, the data you used, how you cleaned and analysed it, what you found, and what you'd do next. Lead with the insight, not the tooling. Know your own project cold — expect follow-ups on every decision you made.
Don't Forget: Questions to Ask Them
Interviews end with "do you have any questions for us?" — and "no" is the wrong answer. Thoughtful questions show genuine interest and help you judge the role:
- What does a typical project look like for an analyst on your team?
- What tools and data sources does the team work with day to day?
- How is success measured in this role in the first six months?
- What does progression look like for a junior analyst here?
Turn Interview Prep Into Real Confidence
You can rehearse answers all day, but nothing beats having genuinely done the work. Real projects give you real stories for the behavioural and walkthrough questions — and verified skills give you certainty on the technical ones. Here's how CadetX helps you get there:
Virtual Work Experience
Get real, company-sourced projects to talk about in your walkthrough and behavioural answers. Free, remote, 3 months, with a portfolio and certificate.
Apply Free →CX Learn
Sharpen and prove your SQL before the technical round — pass a CX Learn exam and earn a free, verified certificate to back it up.
Get Certified Free →Data & AI Launchpad
Want structured interview prep and placement support all the way to hired? Zero to Hired takes you the full distance.
Start Training →Frequently Asked Questions
How do I prepare for a data analyst interview with no experience?
Focus on three things: be able to explain your portfolio projects end to end, practise core SQL and Excel questions out loud, and prepare STAR-structured stories for behavioural questions. Since you have no job history, your projects are your evidence, so know them cold and lead with the business insight, not the tools.
What questions are asked in a data analyst interview?
Data analyst interviews usually mix four types: technical questions on SQL, Excel and visualisation tools; analytical or case questions that test how you approach a problem; behavioural questions about how you have worked in the past; and a walkthrough of a project from your portfolio.
What is the STAR method for interview answers?
STAR stands for Situation, Task, Action, Result. It is a structure for answering behavioural questions: set the scene, explain what you were responsible for, describe what you did, and state the outcome. It keeps your answers focused and evidence-based instead of vague.
What SQL should I know for a data analyst interview?
For most junior roles you should be confident with SELECT statements, WHERE filtering, JOINs, GROUP BY with aggregate functions, ORDER BY, and the difference between WHERE and HAVING. Being able to write and explain a query that answers a business question matters more than memorising syntax.
What questions should I ask at the end of a data analyst interview?
Ask about the tools and data the team works with, what a typical project looks like, how success is measured in the role, and what the path looks like for a junior analyst. Thoughtful questions show genuine interest and help you judge whether the role is right for you.