πHow Data Analyst Interviews Work in 2026
A data analyst interview in India usually has three to four rounds. First comes a screening call or online test with SQL and Excel questions. Then a technical round covering SQL in depth, statistics and a BI tool such as Power BI or Tableau. Many companies add a case round where you get a business problem and a small dataset. The last round is behavioural, often with the hiring manager.
This list follows that order. Read the answers, then practise them out loud and on real data. If you are still building the basics, start with our data analyst roadmap, and for deeper SQL practice see SQL interview questions for data analysts.
πExcel Questions
1. What is the difference between VLOOKUP and XLOOKUP?
VLOOKUP searches only the first column of a range and returns a value from a column to the right, by column number. XLOOKUP searches any column, returns from any column (left or right), defaults to exact match and handles "not found" with a built-in argument. XLOOKUP is the modern default; VLOOKUP still appears in older files, so know both.2. How do you remove duplicates in Excel?
Use Data, Remove Duplicates for a quick clean, choosing which columns define a duplicate. To keep the original data, use the UNIQUE function or Power Query, which records the step so it repeats on the next refresh.3. What is a pivot table and when do you use one?
A pivot table summarises large data by grouping rows and aggregating values, for example total sales by region and month. Use it for fast exploration and quick reports. Add slicers to filter it interactively.4. What is the difference between SUMIF and SUMIFS?
SUMIF takes one condition. SUMIFS takes several, for example sales where region is South and month is March. SUMIFS puts the sum range first, which is a common interview trap.5. How do you handle missing values in Excel?
First find them with filters or COUNTBLANK. Then decide based on meaning: remove the row, fill with a sensible default, or flag it in a helper column. Always note the choice, because it changes the result.6. What is conditional formatting used for?
It highlights cells that meet a rule, such as values above target or duplicates, so patterns jump out without extra charts.πSQL Questions
7. What is the difference between WHERE and HAVING?
WHERE filters rows before grouping. HAVING filters groups after aggregation. "Customers with more than five orders" needs HAVING COUNT(*) > 5.8. Explain INNER, LEFT, RIGHT and FULL joins.
INNER returns only matching rows from both tables. LEFT returns all rows from the left table plus matches, with NULLs where there is no match. RIGHT is the mirror. FULL returns all rows from both sides. Analysts use LEFT joins most, for example all customers with their orders, including customers with none.9. How do you find the second highest salary?
Use a window function: rank salaries with DENSE_RANK() OVER (ORDER BY salary DESC) and keep rank 2. It handles ties correctly, which a simple LIMIT with OFFSET does not.10. What are window functions?
Functions that calculate across a set of rows related to the current row without collapsing them, such as ROW_NUMBER, RANK, LAG, LEAD and running totals with SUM() OVER. They power most "top N per group" and "month over month" questions.11. What is a CTE and why use it?
A Common Table Expression, written with WITH, names a subquery so you can build a query step by step. It makes long analytical queries readable and easier to debug.12. How do you calculate month over month growth in SQL?
Aggregate revenue by month, then use LAG(revenue) OVER (ORDER BY month) to get the previous month, and compute (current minus previous) divided by previous.13. What is the difference between COUNT(*), COUNT(column) and COUNT(DISTINCT column)?
COUNT(*) counts all rows. COUNT(column) skips NULLs in that column. COUNT(DISTINCT column) counts unique non-null values.14. How would you find duplicate records?
Group by the columns that should be unique and keep groups with HAVING COUNT(*) > 1. To delete extras, number rows with ROW_NUMBER() partitioned by those columns and remove rows above 1.πStatistics Questions
15. Mean, median or mode: which do you report?
Mean for roughly symmetric data, median when there are outliers or skew (salaries, order values), mode for categories. Reporting the median order value is usually safer than the mean.16. What is standard deviation?
A measure of how spread out values are around the mean. Low means values cluster closely; high means they vary widely.17. What is a p-value?
The probability of seeing a result at least as extreme as yours if there were truly no effect. A small p-value, often below 0.05, suggests the effect is unlikely to be chance alone. It does not measure how big or important the effect is.18. Correlation versus causation?
Correlation means two variables move together. Causation means one changes the other. Ice cream sales and drowning both rise in summer, yet neither causes the other. Only experiments or careful designs can support causal claims.19. How does an A/B test work?
Split users randomly into a control and a variant, change one thing, run until you reach the planned sample size, and compare a predefined metric with a significance test. Decide the metric and sample size before starting.20. What is an outlier and how do you treat it?
A value far from the rest, often found with the IQR rule or z-scores. Check whether it is an error (fix or remove) or a real event (keep it and maybe report it separately).πPower BI and Tableau Questions
21. What is a star schema?
A data model with a central fact table (sales, transactions) linked to dimension tables (date, product, customer). It keeps BI reports fast and DAX simple.22. Measure versus calculated column in Power BI?
A calculated column is computed row by row at refresh and stored. A measure is computed at query time based on the filters in the visual. Use measures for aggregations like total sales; use columns only when you need a value per row, for slicing.23. What does CALCULATE do?
It evaluates an expression in a modified filter context, for example total sales for the previous year or for one region regardless of slicers. It is the most important DAX function. See Power BI interview questions for more.24. What is row level security?
Rules that restrict which rows a user can see, such as a regional manager seeing only their region, defined with DAX filters on roles.25. Power BI or Tableau?
Both build interactive dashboards. Power BI integrates tightly with Excel and Microsoft 365 and is common in Indian enterprises; Tableau is strong in visual flexibility. Our Power BI vs Tableau guide compares them in detail.26. What makes a good dashboard?
One clear purpose, the key numbers at the top, consistent colours, few chart types, filters that matter, and labels a business user understands without explanation.πPython Basics for Analysts
27. Why do analysts use Python?
For data that is too large or messy for Excel, for repeatable cleaning, and for analysis that needs statistics or automation. pandas is the main library.28. What is a pandas DataFrame?
A table of rows and labelled columns in Python, similar to a sheet or SQL table, with methods to filter, group, join and reshape data. Our pandas tutorial walks through it.29. How do you handle missing values in pandas?
Find them with isna().sum(), then drop with dropna() or fill with fillna(), choosing per column based on meaning.30. groupby in pandas?
It splits data into groups, applies an aggregation, and combines the results, like GROUP BY in SQL: df.groupby("region")["sales"].sum().πCase and Scenario Questions
31. Sales dropped 15 percent last month. How do you investigate?
Confirm the data is correct first. Then break the drop down by region, product, channel and customer segment to find where it is concentrated. Check external factors (seasonality, holidays, pricing changes, outages). Finish with a clear hypothesis and the data you would need to confirm it.32. How would you measure the success of a new feature?
Define the goal, pick one primary metric (such as conversion rate) and a few guardrail metrics (such as refunds or load time), compare against a control group or a before period, and report the result with its uncertainty.33. Which metrics would you track for a food delivery app?
Orders per day, active users, average order value, delivery time, cancellation rate, repeat rate and customer acquisition cost. Group them into growth, operations and retention.34. How do you explain a finding to a non-technical manager?
Lead with the answer and the business impact, show one simple chart, then the reason, then the recommendation. Keep method details for questions.35. How do you prioritise many report requests?
By business impact and urgency, agreeing priorities with stakeholders, and automating repeat requests so they stop needing your time.πBehavioural Questions
36. Tell me about a time you found an error in data.
Use STAR: the situation, the task, what you did to trace the root cause, and the result, including how you prevented it from happening again.37. Describe a project you are proud of.
Pick one portfolio or work project, explain the business question, the data, your method, the result in numbers and what you learned.38. How do you handle a stakeholder who disagrees with your analysis?
Listen to their reasoning, recheck the data and assumptions together, and separate what the data shows from opinion. Offer another cut of the data if it helps.39. How do you keep your skills current?
Mention specific habits: practising SQL problems, building dashboards on public datasets, following tool release notes and reading analytics blogs.40. Why do you want to be a data analyst?
Connect your interest to a real example: a time you used data to answer a question, and the kind of business problems you want to work on.πHow to Prepare in the Last Two Weeks
πFAQ
Is SQL or Excel more important for a data analyst interview?
SQL is tested in almost every data analyst interview and usually carries the most weight. Excel is expected too, especially pivot tables and lookups, but SQL is where candidates most often fail.Do freshers get asked case questions?
Yes, simpler ones. Interviewers want to see structured thinking more than a perfect answer, so talk through your approach step by step.How many projects should I show?
Two strong, well explained projects beat five shallow ones. Our guide on becoming a data analyst without experience covers which projects work best.πLearn It Live
Prakalpana's Data Analytics course covers Excel, SQL, statistics, Power BI, Tableau and Python, taught live online with an optional 1-on-1 track, real projects and mock interviews. Explore the full data analytics learning path, then WhatsApp or call +91 9180377502 for a free demo.