Data Analyst Interview Questions in India (2026)

Data Analyst roles in India are in strong demand heading into 2026. Knok's jobradar tracked 319 active openings as of July 2026, spread across Bangalore (41), Delhi (22), Mumbai (19), Hyderabad (14), Pune (10), and Chennai (5). The highest concentration of jobs is in Bangalore and Delhi NCR, making those cities the primary targets if you are open to relocation.

Interviews typically run two to four rounds: a take-home assignment (SQL problem or a case study), a technical panel with the hiring team, and a final HR or culture-fit conversation. Some firms add a presentation round for senior roles. A two-to-four-week focused preparation period is realistic for most candidates.

Salary bands from those active listings give a useful anchor before you negotiate:

Actual offers depend on company size, domain, and negotiation. Cross-check with Glassdoor or levels.fyi to see what candidates at specific companies commonly report.

1. Walk me through an ML model you shipped to production. What was the business impact?
2. How do you handle imbalanced datasets in a fraud-detection use case?
3. Explain bias-variance trade-off with an example from your work.
4. How would you design an A/B test for a recommendation model?
5. SQL vs Python for feature engineering, when do you use each?
6. Tell me about a time your model performed poorly post-deployment. What did you do?

Q: Write a SQL query to find the second-highest salary in an employee table.

A strong answer demonstrates edge-case awareness, not just the happy path. One clean approach:

`sql
SELECT MAX(salary) AS second_highest
FROM employees
WHERE salary < (SELECT MAX(salary) FROM employees);
`

Mention that DENSE_RANK handles ties more cleanly and generalises to the Nth highest salary. Showing that you think about duplicate salaries and empty tables signals seniority to the interviewer.

Q: A key metric dropped last week. How do you investigate?

Structure your answer before jumping to hypotheses.

'First, I validate the data itself: check for a pipeline failure, a tracking bug, or a metric definition change. Once the data is confirmed clean, I segment the metric by date, region, device, and user cohort to find where the drop is concentrated. Then I list hypotheses based on what changed around that time: a product release, a campaign pause, or seasonality. I test each one with data before presenting a conclusion to the stakeholder.'

Interviewers reward structured thinking over a lucky guess at the root cause.

Q (Behavioural): Tell me about a time your analysis changed a business decision.

Use STAR format.

Situation: 'My team was planning to double the budget on a paid acquisition channel that had shown strong volume for two months.'

Task: 'I was asked to validate the channel ROI before the budget call.'

Action: 'When I joined the acquisition cohorts to long-term retention data, I found that users from this channel churned at a significantly higher rate than organic users. Short-term revenue looked healthy, but three-month LTV was well below our payback threshold.'

Result: 'The team redirected budget to a lower-volume but higher-retention channel. The decision was validated at the next quarterly review when blended LTV improved.'

Keep your example specific and lead with numbers where your company policy allows you to share them.

STAR for behavioural questions: 20% situation, 10% task, 50% action, 20% result with numbers.

CIRCLES for product cases (PM): Comprehend, Identify customer, Report needs, Cut through prioritisation, List solutions, Evaluate trade-offs, Summarise recommendation.

For system design (engineering): clarify scale (DAU, QPS), draw high-level boxes, deep-dive one component, discuss failure modes and monitoring.

Signals that move Data Analyst candidates forward in India:
- Ownership of outcomes, not tasks
- Comfort with ambiguity and incomplete data
- Collaboration with cross-functional partners
- Understanding of India-specific constraints (UPI, logistics, multilingual users, price sensitivity)
- Realistic salary expectations aligned with level

A two-to-four-week plan works for most candidates preparing from a working baseline.

Week 1: SQL and Statistics Foundations

Spend the first week solidifying SQL. Work through practice problems on LeetCode SQL, HackerRank, or Mode Analytics tutorials (all publicly available). Prioritise window functions and multi-table queries: these are the most commonly tested areas. Also revise statistics basics: distributions, outliers, and A/B testing concepts. Write brief notes in your own words rather than copying definitions.

Week 2: Tools and Case Practice

Build or revisit one small project in Python or Excel that answers a real business question. Prepare two or three dashboard examples you can describe confidently under interview pressure. Start practising case questions out loud: given a metric drop or a business problem, structure your answer verbally before writing anything. This mirrors the actual interview format more closely than silent practice.

Week 3: Behavioural Stories and Mock Interviews

Write out four to six STAR stories covering: working with messy data, influencing a decision with analysis, handling a stakeholder disagreement, and delivering under a deadline. Do at least one timed mock interview with a peer, or record yourself answering out loud. Watching the recording is uncomfortable but it reveals filler words and unclear explanations you cannot catch in the moment.

Week 4: Company Research and Refinement

For each company you interview with, read recent news, understand their core product or business model, and prepare genuine questions for the interviewer. Revisit any weak areas identified in mock sessions. Keep SQL sharp by solving one problem each morning.

If you are short on time, prioritise SQL and one strong behavioural story. Those two cover the majority of early-round filters. While you prepare, knok checks 150+ job sites nightly, applies to roles that match your resume, and messages HR for you, so your applications keep moving even when you are deep in practice mode.

• Rambling without a clear result metric
• Badmouthing previous employers
• Quoting global salary data without India context
• Ignoring the 'why this company' question
• Over-indexing on frameworks without showing real shipped work

Candidates report that Data Analyst interviews in India follow a predictable arc: SQL and data fundamentals first, statistical thinking second, and behavioural questions in later rounds. The questions below are commonly asked across product, fintech, e-commerce, and consulting firms.

SQL and Data Fundamentals

1. Write a query to find the second-highest salary in an employee table.
2. What is the difference between a LEFT JOIN and an INNER JOIN? Give a practical example.
3. How do you handle NULL values in a dataset before running an analysis?
4. What are window functions? Explain ROW_NUMBER, RANK, and DENSE_RANK with a use case.
5. What is normalization? When would you deliberately denormalize a table?

Statistics and Analytical Thinking

1. What is the difference between correlation and causation? Give a business example.
2. How do you detect and treat outliers in a dataset?
3. A key product metric dropped sharply last week. Walk me through your root-cause investigation.
4. What is A/B testing, and how do you decide whether a result is statistically significant?

Tools, Dashboards, and Business Scenarios

1. How have you used Python or Excel to automate a repetitive reporting task?
2. A stakeholder disagrees with your analysis. How do you handle it?
3. Describe a dashboard you built. What business decisions did it drive?

Behavioural (commonly added in later rounds)

Candidates also report one or two open-ended behavioural questions. The most frequently cited: 'Tell me about a time you worked with messy or incomplete data' and 'Describe a situation where your analysis changed a business decision.' Use the STAR format (Situation, Task, Action, Result) for both.

SQL (non-negotiable for every interview)

Every Data Analyst interview in India includes at least one live SQL problem. Focus on JOINs (all four types), GROUP BY with HAVING, subqueries, CTEs, and window functions: ROW_NUMBER, RANK, DENSE_RANK, LAG, and LEAD. Practice writing queries from scratch without auto-complete, because most live rounds use a shared screen or a basic browser-based editor.

Python or Excel for Data Wrangling

For Python: pandas (groupby, merge, pivot_table), basic matplotlib or seaborn for visualisation, and simple automation scripts. For Excel: pivot tables, XLOOKUP, and conditional formatting. Entry and mid-level roles often accept either; senior roles and product companies increasingly expect Python.

Statistics Fundamentals

Mean, median, mode, variance, and standard deviation: know when to use each. Distributions (normal, skewed, bimodal). Hypothesis testing: p-value, confidence intervals, Type I and Type II errors, and statistical power. A/B testing design and interpretation. You do not need graduate-level mathematics. You need to explain these concepts clearly to a non-technical stakeholder, which is what interviewers actively test.

Data Visualisation and Storytelling

Familiarity with at least one BI tool: Tableau, Power BI, and Looker are commonly cited in Indian job listings. More important than tool mastery is choosing the right chart type and explaining what the data means for a business decision. Prepare to narrate the story behind one end-to-end dashboard you have built.

Business and Domain Metrics

Know the core metrics for the industry you are targeting. For e-commerce: GMV, conversion rate, cart abandonment. For fintech: NPA, CAC, LTV. For SaaS: MRR, churn rate, DAU/MAU. Interviewers test whether you can connect data to outcomes, not just run queries.

Communication and Stakeholder Management

Translating a complex finding into a one-sentence business recommendation is a skill interviewers probe directly. Practice narrating your work out loud, not just building it.

Jumping to the answer without clarifying the question

Interviewers often leave questions deliberately vague. Before writing a query or diving into analysis, ask one clarifying question. This signals seniority and prevents solving the wrong problem entirely.

Writing SQL that only works on the happy path

Candidates who forget about NULLs, duplicate values, or empty result sets lose marks. Always mention edge cases out loud, even if you do not write the full solution for each one.

Memorising answers instead of understanding concepts

Interviewers follow up. If you recite a definition of p-value but cannot explain what it means for a specific business scenario, the next question will expose the gap. Build genuine understanding, not flashcard responses.

Skipping the business context in technical answers

A technically correct SQL answer that ignores the business question it was meant to serve is a yellow flag. Always connect your output to a decision or action the stakeholder can actually take.

Underselling communication skills

Data Analysts spend a large part of their time explaining findings to people who did not ask for the analysis in a technical format. Show that you can translate a query result into a plain recommendation. Interviewers look for this explicitly, especially at mid and senior levels.

Not preparing questions to ask the interviewer

Candidates who have no questions signal low motivation. Prepare two genuine questions about the team's data stack, the metrics they care about most, or how the role has grown over the past year. Avoid raising salary in the first technical round.

Ignoring the take-home assignment brief

If the brief says 'summarise your findings in one slide,' sending ten slides is a red flag. Following the instructions precisely signals the attention to detail that is central to this role.