In this guide
Role-specific resume and interview guides
Data analyst interviews: SQL tests, case questions and presenting findings
How to prepare for data analyst interviews: practical SQL and spreadsheet tests, metric investigation questions, statistics basics and presenting findings.
Practical tests: what to practise
| Task | Skills tested | Practice focus |
|---|---|---|
| SQL queries | Joins, grouping, filtering, window functions | Write queries without autocomplete; check row counts after each join |
| Spreadsheet task | Lookups, pivot tables, cleaning | Explain each step and keep the raw data untouched |
| Take-home analysis | End-to-end reasoning | A short written summary of findings and limits |
| Dashboard critique | Visual judgement | Say what question each chart answers and what is missing |
Talk while you work if the test is live. Saying 'I'll check for duplicate customer IDs before counting' shows the habits employers want, even if you run short of time.
Metric investigation questions
A common case question is a variation of 'weekly sales dropped by ten per cent; how would you investigate?'. Interviewers are testing whether you check the data before the business, and whether you break the problem down systematically.
Illustrative example
A structure for a metric-drop question
- Check the data: did tracking, definitions or data loads change? Is the drop real?
- Size it: when did it start, how large is it, is it outside normal variation?
- Segment it: by region, channel, product, customer type and device.
- Look for causes: pricing, stock, campaigns, competitors, seasonality, outages.
- Recommend: what to check next, who to ask and how you would confirm a cause.
Statistics and judgement questions
Expect conceptual rather than mathematical questions in most analyst roles: the difference between correlation and causation, why a small sample is unreliable, how you would read an A/B test result, or when an average misleads. Prepare a plain-English answer and an example from your own work for each.
Presenting a past analysis
- The question and who asked it.
- The data you used and the problems you had to fix.
- What you found, in one sentence.
- The decision it informed, and any limitation you flagged.
Before an analyst interview
- I have practised joins and aggregations without help from autocomplete.
- I can explain a past analysis in under two minutes.
- I have a structure ready for a metric investigation question.
- I can explain one statistics concept with an example.
- I know which tools the team uses and which I would need to learn.
Common questions
Do data analyst interviews always include a SQL test?
Not always, but many do. Ask the recruiter what the practical stage involves and which tools or dialect you can use.
What if the test uses a tool I have not used?
Say so honestly, explain how you would do it in a tool you know and how quickly you have learned tools before.
Put it into practice
See how Luceria helps with thisDrafted with AI assistance and checked against Luceria’s editorial standards. Examples are illustrative, not real people’s results. Spotted something out of date? Tell us.