Data Analyst Interview Questions & Answers

Tell us about an analysis that changed a decision or had real impact.

At my last company we assumed mobile conversion was lower because of pricing, and there was pressure to discount. I dug into a year of funnel data, segmented by device and step, and found the real drop-off was on the shipping-address screen on Android specifically — not pricing at all. I cross-referenced it with error logs and traced it to an autofill bug. I put the finding into a one-page deck with a single clear chart and walked product and engineering through it. They fixed the bug instead of discounting, and mobile conversion recovered about 8%. The lesson I always carry is to segment before trusting an aggregate — the headline number hid the real story.

How would you explain a complex finding to a non-technical stakeholder?

I start from the decision they need to make, not the methodology. I lead with the 'so what' — the one sentence that tells them what to do — then support it with a single clear visual rather than a wall of numbers. I avoid jargon, use plain comparisons, and keep the technical detail in an appendix for anyone who wants it. For example, instead of 'the regression coefficient was significant,' I'd say 'customers who do X are roughly twice as likely to churn, so it's worth targeting them.' Checking they've actually understood — and inviting questions — matters more than showing how clever the analysis was.