Data Analyst Interview Questions With Worked Examples

Last updated: 21 September 2026 · By the Asuraa Team

What are the common data analyst interview questions?

Data analyst interview questions usually cover your background, how you work with data, your tools, and some statistics or business judgement. Coursera's list of analyst interview questions, updated 27 June 2026, groups them into general and behavioural questions, data-analysis process questions and technical skills questions.

The table maps the categories to what interviewers are checking. The examples come from Coursera's list or from our own practice set.

How should you answer SQL questions? (worked examples)

Say your logic first, then write the query, then check the result on a small example. Coursera lists SQL question types including writing a query with JOIN and COUNT, describing what an existing query does, and debugging one.

The examples below use an invented practice dataset, and we ran each query to check the output. The orders table has order_id, customer_id, city and amount, and the customers table has customer_id and name.

Question 1: Find customers whose orders of Rs 500 or more add up to more than Rs 1,000.

SELECT customer_id, SUM(amount) AS total
FROM orders
WHERE amount >= 500
GROUP BY customer_id
HAVING SUM(amount) > 1000;

On our practice data this returns customer 1 with 2,000 and customer 3 with 2,500. The PostgreSQL documentation explains why both clauses appear: WHERE selects input rows before groups and aggregates are computed, while HAVING selects group rows after they are computed.

Question 2: Which customers have never placed an order?

SELECT c.name
FROM customers c
LEFT JOIN orders o ON o.customer_id = c.customer_id
WHERE o.order_id IS NULL;

A LEFT JOIN keeps every customer, and customers with no matching order get a NULL order id. On our data it returns Karan, the only customer with no orders.

Question 3: Show the largest order in each city.

SELECT city, order_id, amount
FROM (
  SELECT city, order_id, amount,
         ROW_NUMBER() OVER (PARTITION BY city ORDER BY amount DESC) AS rn
  FROM orders
) t
WHERE rn = 1;

This numbers the orders within each city from largest to smallest and keeps the first. On our data it returns order 104 (2,500) for Delhi and order 102 (1,200) for Pune. For more practice, see our post on SQL interview questions.

How should you answer Excel questions?

Explain what the function does, give a formula, and mention its limits. Coursera says Excel questions often cover VLOOKUP and its limitations, pivot tables, removing duplicates, and INDEX and MATCH.

Microsoft's VLOOKUP page gives the syntax =VLOOKUP(lookup_value, table_array, col_index_num, [range_lookup]). It says the value you look up must be in the first column of the range, that FALSE gives an exact match and returns #N/A if none exists, and that leaving the last argument out defaults to approximate match. It also suggests trying XLOOKUP, which works in any direction.

A strong answer to "How would you use VLOOKUP?" says: "To fetch a price by product code, I would use exact match, so a missing code shows #N/A and not a wrong price. If the code column is not first, I would use XLOOKUP or INDEX and MATCH."

How should you answer statistics questions?

Explain the idea in plain words, then say when it can mislead. Coursera says to be familiar with the mean, standard deviation, variance, regression, and descriptive and inferential statistics.

Mean or median? Take eight delivery times in minutes: 20, 22, 25, 26, 28, 30, 32 and 400. The mean is about 72.9, but the median is 27, because the one extreme value of 400 pulls the mean up. Say that the median describes a typical delivery better here, and that you would also investigate why one delivery took 400 minutes.

What does a p-value tell you? The American Statistical Association's statement on p-values says p-values can indicate how incompatible the data are with a specified statistical model. It also says they do not measure the probability that a hypothesis is true, and do not indicate effect size or importance. In an interview, add that a small p-value does not mean the result matters to the business.

How do you handle a business case question? (an illustration)

Clarify the question, split the problem, check the data, then recommend. This structure is our own, and the scenario is invented: "Orders fell 10% this week. What would you do?"

1. Clarify. Ask about the time frame, the definition of an order and whether tracking changed.
2. Check the data first. Look for missing days, duplicates or a broken tag before blaming customers.
3. Segment. Split by city, product category, channel and new versus returning customers.
4. Form hypotheses. Compare the segments, and test the likeliest cause, such as a stock-out or a campaign ending.
5. Recommend. State the finding, the confidence you have and the next action.

Coursera also lists guesstimate questions, where you think aloud about the data you need and how you would calculate an estimate. The same habit of stating assumptions works there.

How should you answer behavioural and process questions?

Use a short story with a result. Coursera recommends the STAR framework (situation, task, action, result) and advises explaining technical ideas to non-technical audiences.

For "What is your process for cleaning data?", describe how you handle missing values, duplicates, structural errors and outliers, and how you record each change. For "What is the largest dataset you have worked with?", Coursera says portfolio projects and coursework are valid examples. Our guides on how to prepare for a job interview in India and HR interview questions for freshers cover the general rounds.

How should you prepare in a week?

Spend a week on the skills interviewers test, using your own projects as material. This plan is our editorial suggestion.

1. Day 1: Rewrite your two best projects as two-minute stories with the question, method and result.
2. Day 2: Practise five SQL queries covering JOIN, GROUP BY, HAVING and a window function.
3. Day 3: Practise VLOOKUP, a pivot table and duplicate removal in Excel.
4. Day 4: Explain mean, median, variance and p-values aloud in plain language.
5. Day 5: Work through two business cases with the structure above.
6. Day 6: Practise "tell me about yourself" and your questions for the interviewer.
7. Day 7: Do a mock interview with a friend and note where you rambled.

Projects give you material for almost every answer, so see our post on data analyst portfolio projects if you need one, and revise the data analyst skills you plan to claim.

What do most guides on data analyst interview questions get wrong?

Most guides are long lists of questions with a one-line answer each. These are the gaps.

• They reward memorising. Interviewers often ask you to write a query or reason aloud, so practise doing, not reciting.
• They skip the data-checking step. In cases, a strong answer first asks whether the data itself is wrong.
• They ignore your own projects. Many questions are answered best with an example from your work.
• They treat statistics as definitions. Interviewers care whether you know when a number can mislead.

FAQ

What are the most common data analyst interview questions?

Common questions cover tell me about yourself, your process for cleaning data, SQL queries with joins and grouping, Excel functions such as VLOOKUP, basic statistics, and a business case. Coursera groups them into general, process and technical questions. Prepare an example from your own project for each.

How do I prepare for a data analyst interview as a fresher?

Rehearse two projects as short stories, practise SQL and Excel tasks by hand, and explain statistics aloud in plain words. Coursera recommends the STAR method and building a portfolio. Spend a week on this, then do a mock interview and note where you were unclear.

What SQL questions are asked in data analyst interviews?

Coursera lists five types: writing a query with JOIN and COUNT, describing an existing query, changing records, debugging a query and defining terms such as primary key or union. Practise joins, grouping with HAVING, and window functions, and say your logic before you type.

What is the difference between WHERE and HAVING in SQL?

The PostgreSQL documentation says WHERE selects input rows before groups and aggregates are computed, while HAVING selects group rows after they are computed. Use WHERE to filter rows and HAVING to filter aggregated results, such as customers whose total spend exceeds a limit.

What Excel questions should a data analyst expect?

Expect VLOOKUP and its limits, pivot tables, removing duplicates, INDEX and MATCH, and the difference between functions and formulas, according to Coursera. Microsoft notes that VLOOKUP needs the lookup value in the first column and suggests XLOOKUP as a more flexible option.

How do I answer a case question in a data analyst interview?

Clarify the question, check the data quality, split the problem into segments, form and test hypotheses, then give a recommendation with your confidence. Say your assumptions aloud. This structure is our own, and interviewers mostly want to see clear, organised thinking.

Final thoughts

Interviewers want to see you reason with data and explain it clearly. Work through real examples aloud, use your own projects as evidence, and keep the checks in your answers: is the data right, and could this number mislead?

If your resume is ready but the interviews are not coming, check it with the AI resume reviewer on asuraa.in, which reports ATS-compatibility feedback and keyword gaps against a job description.