Data Analyst Interview Questions (SQL, Dashboards & Case Scenarios)

In a Data Analyst interview, strong candidates prove they can convert raw data into a decision-ready recommendation, not just produce correct queries. Expect questions that test how you define metrics, validate data quality, and explain trade-offs using real KPIs such as conversion rate, churn, CAC or margin. Interviewers commonly look for clarity around calculation logic—for example, whether conversion uses unique users versus sessions or whether churn uses a snapshot-based denominator. Tools and workflows often come up naturally, such as SQL window functions for time-series analysis, Power BI for dashboard consistency, and dbt-style modelling for repeatable transformations.

Rehearse SQL that goes beyond basic SELECT statements, including CTEs for readability, window functions for rolling metrics, and GROUP BY + HAVING for threshold-based logic. A typical challenge is to compute rolling conversion or compare basket values across time windows while avoiding leakage between periods. Interviewers also test whether you understand join cardinality—incorrect joins can inflate totals and quietly damage KPIs shown in Power BI or Tableau. If you can explain how you protect metric integrity (e.g., ensuring the fact grain is correct and validating denominator counts) you’ll stand out. Where relevant, reference indexing or partitioning strategies as part of performance thinking, such as indexing (customer_id, purchase_date) for fast time filters.

For executive stakeholders, a dashboard must answer three questions quickly: what changed, why it changed, and what to do next. Build your approach around KPI hierarchies—top-level trends (e.g., revenue, churn, NPS, MRR or CAC) with drill-down by product, channel and geography—so leaders can move from overview to root cause. Use time intelligence such as 12-month rolling views and YoY comparisons, and include alerting logic when thresholds breach (for example, margin falling more than 3% over a defined baseline). When discussing implementation, mention how you’d set up dynamic filters, scheduled refresh, and consistent measure definitions in Power BI (DAX) or Tableau calculated fields. The best answers tie chart choices directly to decision rhythms, such as weekly review meetings or monthly board reporting, with measurable outcomes like “identified an 8% margin decline in one segment within one meeting.”

In cases, you may be told “the churn is wrong” or “the dashboard looks off”, so your job is to debug the metric definition and the data pipeline. Start with a metric reconciliation plan: confirm numerator/denominator logic, validate join keys, check for late-arriving data, and segment the issue to isolate whether it’s a definition change, data quality breach, or pipeline regression. Then shift into explanation and action by identifying likely drivers—such as onboarding issues affecting retention or channel changes impacting conversion—backed by evidence. When data is incomplete or deadlines are tight, deliver a V1 with explicit assumptions, quantify known gaps, and propose a timeline for a “v2” once additional data lands. This demonstrates both analytical rigour and professional judgement, which is central to effective stakeholder management in analytics teams.