Mock Interview for Data Analysts: SQL, Excel, Statistics and Case Questions with a Report

Data analyst interviews in India have settled into a recognisable shape: a SQL test or live query, a few Excel or BI-tool questions, some statistics explained in plain words, and one open business problem that checks whether you can think before you query. Candidates who know all four areas still fail the round, because the questions come with follow-ups and the follow-ups are spoken, not typed.

This page lists what analyst interviewers actually ask at the fresher to three-year level, how a MockMate Practice session runs for the role, what the report tells you, and a seven-day plan.

Who this is for

• Graduates from any stream who have learned SQL, Excel and one of Python, Power BI or Tableau, and are applying for analyst roles at IT services firms, consulting, e-commerce, fintech and startups.
• Working analysts with one to three years who want to move from reporting into a role with more modelling or product analytics.
• MBA and commerce graduates applying for business analyst and MIS roles where the interview is half analysis and half communication.

For the full fresher question bank with sample answers, see data analyst interview questions for freshers.

What the interviewer actually asks

1. "Tell me about a dataset you analysed end to end." Strong answer: the business question, where the data came from, what cleaning it needed, the method, the insight, and what someone did differently because of it.
2. "Write a query for the top three products by revenue in each category." Strong answer: a window function (RANK or ROW_NUMBER with PARTITION BY), and a sentence on how ties are handled.
3. "What is the difference between INNER, LEFT and FULL OUTER JOIN, and when does a LEFT JOIN inflate your counts?" Strong answer: the definitions in one line each, then the duplicate-rows-on-the-right problem and the aggregation fix.
4. "How do you handle missing values?" Strong answer: it depends on why they are missing; drop, impute or flag; and check how the choice moves the result.
5. "Explain a p-value or an A/B test result to a manager who is not technical." Strong answer: no jargon, one sentence on what the number means for the decision, and what you would recommend.
6. "VLOOKUP versus INDEX-MATCH versus XLOOKUP; when do you use a pivot table?" Strong answer: when each applies, one limitation of VLOOKUP, and a pivot example from your own work.
7. "Sales dropped 15 percent last month. How would you investigate?" Strong answer: check the data pipeline first, then segment by region, product, channel and time, form two or three hypotheses, and say which you would test first and why.
8. "Which chart would you use, and why?" Strong answer: bars for comparison, lines for trends, histograms for distribution, and why a pie chart with twelve slices is wrong.
9. "Mean versus median: when does the mean mislead?" Strong answer: skewed data and outliers, with the salary example.
10. "In pandas or Power BI, how would you..." (groupby and merge, or a DAX measure). Strong answer: know one tool deeply; say clearly which one and show it.

Every one of these has a follow-up. "Top three per category" is followed by "what if two products tie for third?"; "how would you investigate" is followed by "you have one hour, what do you do first?"

How a MockMate Practice session runs for this role

Practice runs in the browser.

1. Attach your resume and the job description. Questions reflect the tools listed on your resume, the seniority you set, and the job description's emphasis (a fintech JD produces different case questions from a retail one). State a focus area such as "SQL and case questions" or "statistics and communication".
2. Pick the round. A technical round includes a code editor, which is where the SQL questions live. An HR-style round covers the project walkthrough, the "explain to a manager" questions and behavioural questions using the STAR method.
3. Adaptive follow-ups. The interviewer persona reacts to your answer. Name a window function and it asks about ties; propose imputing a missing value and it asks how you would check the effect.
4. Ten minutes per free start, which fits four to six questions with follow-ups. Eligible accounts get three free Practice starts, tracked server-side.
5. A report at the end. Details below.

What the report tells you

• Question-level answers with coaching on what a stronger answer would have contained, for example "named the join but did not say what the result set contains".
• Response timing: how long before you started speaking and how long you spoke. Analysts often over-explain; the timing shows it.
• Coaching evidence: the exact phrase or omission behind each coaching note.
• Recurring weaknesses across the session, such as "jumps to a method before stating the business question".
• A recommended next practice: the round type and focus to run next.

A 7-day prep plan

Common mistakes in data analyst rounds

The Practice report tends to flag the same five patterns in analyst candidates.

1. Querying before asking. Given "sales dropped 15 percent", the candidate starts writing SQL. The strong version first asks what "sales" means here, over what period, and whether the pipeline changed.
2. Tool lists instead of depth. A resume with SQL, Python, R, Excel, Tableau, Power BI and SAS invites a hard question in the weakest one. Name the two you know well and be ready to go deep.
3. Statistics explained in jargon. "The p-value was below the alpha threshold" to a sales manager means nothing. "We can be fairly confident the new page performed better, and here is the risk" is the answer.
4. Joins without result sets. Writing a correct LEFT JOIN but not being able to say how many rows it returns or why counts inflated. Always describe the output.
5. Charts chosen by habit. A pie chart for twelve regions, a line chart for categories. Say why the chart fits the question before you draw it.

A sixth, common for freshers: describing a college project as "I analysed the data and found insights" with no business question, no cleaning step and no decision at the end. The end-to-end story in question 1 above is the fix.

Honest limits

• Practice is a browser product; Live is Windows-only. The mock interview needs a modern browser and nothing else. MockMate's Live assistance companion runs only on Windows 10 and 11; there is no Mac, iOS or Android app.
• Live is for permitted settings only. Use it only where the organisation, interviewer or applicable rules permit assistance and disclosure; use Practice when permission is unclear. Take-home assignments and proctored SQL tests are no-assistance settings.
• The persona is not your hiring manager. It does not know the company's data stack. One conversation with a working analyst about their day-to-day is still worth having.
• It does not replace writing queries. The session shows that your window-function answers are shaky; day 2 fixes them.
• The free allowance is three ten-minute starts. After that, credits apply (as of September 2026: Starter ₹2,000 for 2 credits, Growth ₹5,000 for 5, Pro ₹10,000 for 12; one credit is about sixty minutes). See pricing.

If your target role is closer to engineering than analysis, the software engineer mock interview page covers the DSA and design side.