💻Technical Questions
Q1Write a SQL query to find the second-highest salary in a table.
💡Multiple approaches: OFFSET/LIMIT, subquery with MAX, DENSE_RANK(). Show at least 2 methods.
Q2What is the difference between RANK(), DENSE_RANK(), and ROW_NUMBER()?
💡RANK() skips ranks for ties; DENSE_RANK() doesn't skip; ROW_NUMBER() is always unique. Use cases for each.
Q3Explain the difference between WHERE and HAVING clauses.
💡WHERE filters rows before aggregation; HAVING filters groups after aggregation. Cannot use aggregate functions in WHERE.
Q4How would you detect outliers in a dataset?
💡IQR method (1.5 × IQR rule), Z-score (>3 standard deviations), visual methods (box plots, scatter plots), domain-specific thresholds.
Q5What is a cohort analysis and when would you use it?
💡Group users by a common event/time (signup month) and track behavior over time. Used for retention, LTV, engagement trend analysis.
Q6Explain the difference between inner join, left join, and full outer join.
💡Use diagrams mentally. Inner: matching rows only. Left: all from left + matching right. Full outer: all rows from both. Anti-join use case.
Q7What is A/B testing? How do you determine if results are statistically significant?
💡Hypothesis testing, p-value (<0.05), statistical power (80%), sample size calculation, avoiding peeking problem.
🧠Behavioral Questions
B1Tell me about a time your analysis changed a business decision.
💡Structure: what was the question, what data you used, what you found, and the outcome. Quantify the impact.
B2Describe a time you had to work with messy or incomplete data.
💡How you identified gaps, your handling strategy (imputation, exclusion, flagging), and how you communicated data quality limitations to stakeholders.
B3How do you explain a complex analysis to a non-technical stakeholder?
💡Avoid jargon, lead with the insight not the method, use visualizations, anticipate business-focused questions.
🎯Situational Questions
S1DAU dropped 15% last week with no product changes. Walk me through how you'd investigate.
💡Segment by platform, geography, user cohort, feature. Check data pipeline issues first. Then: acquisition drop, engagement drop, churn increase. Each has different fixes.
S2A PM wants to launch a feature without an A/B test because 'we're confident it'll work'. What do you do?
💡Understand their reasoning, propose minimum viable test, estimate sample size needed, offer holdout group as compromise.
Must-Know Topics
- ✓SQL (window functions, CTEs, subqueries)
- ✓Statistics (hypothesis testing, distributions, regression)
- ✓Excel / Google Sheets (pivot tables, VLOOKUP)
- ✓Python or R (pandas, matplotlib)
- ✓A/B Testing & Experimentation
- ✓BI Tools (Tableau, Power BI, Looker)
- ✓Business Metrics (DAU, MAU, LTV, CAC, churn)
- ✓Data Storytelling
Common Interview Mistakes to Avoid
- ✗Diving into analysis without defining the metric to move
- ✗Confusing correlation with causation in business recommendations
- ✗Not checking for data quality issues before drawing conclusions
- ✗Forgetting to segment data (Simpson's paradox)
- ✗Presenting numbers without business context or recommendation
Frequently Asked Questions
Is SQL the most important skill for a data analyst interview?▼
Yes — SQL is tested in almost every data analyst interview. Focus on window functions (RANK, LAG, LEAD, SUM OVER PARTITION), CTEs, self-joins, and complex aggregations. Aim to write clean, readable SQL without looking up syntax.
How much Python/R do I need to know for a data analyst interview?▼
For most analyst roles (not data science), Python proficiency means: pandas for data manipulation, matplotlib/seaborn for visualization, basic statistical tests with scipy. You typically don't need machine learning for analyst interviews.
What business case study types are common in data analyst interviews?▼
Metric deep-dives (why did X drop?), A/B test interpretation, product launch analysis, funnel analysis, cohort retention analysis. Practice structuring your approach before jumping to analysis.
What BI tools should a data analyst know?▼
Tableau and Power BI are the most common in India. Looker is common at US-based product companies. Most companies will teach you their internal tool — knowing one well is sufficient.
How do I answer 'how would you approach this analysis' questions?▼
Structure: (1) Clarify the business question, (2) Define the metric, (3) Identify data sources, (4) List assumptions, (5) Outline the analysis approach, (6) Describe how you'd validate findings, (7) How you'd present results.
Ready for your Data Analyst interview?
Make sure your resume gets you to the interview stage first. Get a free ATS score.
Score My Resume Free →