Data Analyst Interview Questions & Answers: 40+ Questions for 2026

The SQL question looks simple: "Write a query to find the top 3 products by revenue for each region." You have 15 minutes. Your hands are sweating. The interviewer is watching you think out loud.

This is the moment that separates candidates who studied from candidates who prepared — and the difference is worth $20,000-$30,000 in starting salary. Data analyst interviews test three things at once: technical skill, analytical reasoning, and communication under pressure. Most candidates only prepare for one.

The questions themselves are predictable. The way interviewers evaluate your answers is not — and understanding the scoring criteria changes everything about how you prepare.

Brought to you by Careery

This article was researched and written by the Careery team — that helps land higher-paying jobs faster than ever! Learn more about Careery →

Walk into a startup interview prepared for big tech, and you'll waste everyone's time. Walk into big tech prepared for a startup, and you'll get destroyed. Format matters.

The interview process varies dramatically depending on where you're applying. A startup might extend an offer after two conversations. A big tech company might put you through five rounds over a month.

The technical assessment is the gate. Let's start with the area that appears in nearly every data analyst interview: SQL.

SQL appears in nearly every data analyst interview — and the gap between passing and failing is smaller than most people think. One missed edge case, one sloppy NULL handling, and you're out.

SQL is the universal data analyst skill. Every company with a database needs someone who can query it. The questions below cover the most tested patterns, from foundational to advanced.

"Write a query to find the second-highest salary in each department."

This tests window functions — the skill that separates junior SQL from production SQL.

SELECT department, employee_name, salary
FROM (
  SELECT
    department,
    employee_name,
    salary,
    DENSE_RANK() OVER (PARTITION BY department ORDER BY salary DESC) as rank
  FROM employees
) ranked
WHERE rank = 2;

"Given an orders table and a customers table, find customers who placed orders in January but NOT in February."

This tests your ability to combine JOINs with conditional logic.

SELECT DISTINCT c.customer_id, c.customer_name
FROM customers c
JOIN orders o ON c.customer_id = o.customer_id
WHERE EXTRACT(MONTH FROM o.order_date) = 1
  AND EXTRACT(YEAR FROM o.order_date) = 2026
  AND c.customer_id NOT IN (
    SELECT customer_id
    FROM orders
    WHERE EXTRACT(MONTH FROM order_date) = 2
      AND EXTRACT(YEAR FROM order_date) = 2026
  );

"Calculate the 7-day rolling average of daily revenue."

Window functions with frames — a common real-world pattern for trend analysis.

SELECT
  order_date,
  daily_revenue,
  AVG(daily_revenue) OVER (
    ORDER BY order_date
    ROWS BETWEEN 6 PRECEDING AND CURRENT ROW
  ) as rolling_7_day_avg
FROM (
  SELECT
    DATE(order_date) as order_date,
    SUM(amount) as daily_revenue
  FROM orders
  GROUP BY DATE(order_date)
) daily;

"Find the month-over-month revenue growth rate."

LAG function — essential for any time-series analysis.

WITH monthly_revenue AS (
  SELECT
    DATE_TRUNC('month', order_date) as month,
    SUM(amount) as revenue
  FROM orders
  GROUP BY DATE_TRUNC('month', order_date)
)
SELECT
  month,
  revenue,
  LAG(revenue) OVER (ORDER BY month) as prev_month_revenue,
  ROUND(
    (revenue - LAG(revenue) OVER (ORDER BY month)) * 100.0
    / LAG(revenue) OVER (ORDER BY month), 2
  ) as growth_rate_pct
FROM monthly_revenue
ORDER BY month;

SQL gets you through the technical gate. But most data analyst roles require more than just querying databases.

Most candidates over-prepare for SQL and under-prepare for Excel. Then they freeze when asked to find duplicates in a 50,000-row dataset during a live screenshare.

"How would you find duplicate values in a dataset of 50,000 rows?"

Multiple approaches — the answer reveals your Excel depth:

• COUNTIF method: Add a helper column with =COUNTIF(A:A, A2). Any value > 1 is a duplicate. Sort by the helper column to group them.
• Conditional formatting: Select the range → Conditional Formatting → Highlight Cell Rules → Duplicate Values. Quick visual identification.
• Remove Duplicates tool: Data tab → Remove Duplicates. But this destroys data — always work on a copy.

"When would you use INDEX-MATCH instead of VLOOKUP?"

VLOOKUP only searches left to right — the lookup column must be the leftmost column. INDEX-MATCH works in any direction and is more flexible:

• VLOOKUP: =VLOOKUP(lookup_value, table_array, col_index, FALSE) — simple, but breaks if columns are rearranged.
• INDEX-MATCH: =INDEX(return_range, MATCH(lookup_value, lookup_range, 0)) — more robust, handles any column arrangement, and performs better on large datasets.

Excel proves you can work in the tool stakeholders actually use. But the next category tests something harder to fake: whether you understand what the data actually means.

This is where candidates with bootcamp certificates get separated from candidates who actually think like analysts. The questions sound simple. The wrong answers cost offers.

"What is the difference between correlation and causation?"

Correlation means two variables move together. Causation means one variable directly causes the other to change. Ice cream sales and drowning deaths are correlated (both increase in summer) but ice cream doesn't cause drowning — temperature is the confounding variable.

"You run an A/B test and get a p-value of 0.03. What does this mean?"

A p-value of 0.03 means there's a 3% probability of observing results this extreme (or more extreme) if the null hypothesis were true — that is, if there were actually no difference between the two groups.

"Revenue is up 15% month-over-month. What questions would you ask before reporting this to the executive team?"

This is the question that separates analysts from query-runners:

• Is the increase driven by a single outlier (one large order) or broad-based growth?
• Did anything change externally — a seasonal pattern, a marketing campaign, a pricing change?
• Is the comparison period representative? (Was last month unusually low?)
• Are we looking at the same customer cohort, or did we add new accounts?
• Is the 15% statistically significant given normal revenue variance?

Understanding statistics is one thing. Applying it under pressure — with a vague prompt and a ticking clock — is where the real test begins.

The case study is the closest thing to doing the actual job during an interview. It's also where the highest-scoring candidates on technical rounds suddenly fall apart.

The Framework for Any Case Study

Every data analyst case study follows the same underlying structure, regardless of the specific question:

You've survived the technical rounds. Now comes the part that decides whether you're a culture fit — and whether you can communicate what you found.

Technical skills get you to the final round. Behavioral answers determine who gets the offer — and most data analysts treat this section as an afterthought.

Behavioral questions appear in every data analyst interview. The answers that stand out use the STAR format with data-specific details.

Example Behavioral Questions

Strong answer structure: Describe the original direction the team was heading → explain the analysis you ran → show how the data contradicted the assumption → describe the new decision and its impact.

This question tests intellectual honesty. The best answers acknowledge the mistake openly, explain what caused it (data quality issue, flawed assumption, missing context), and describe the process change you implemented to prevent it from happening again.

The answer interviewers want: "Start with the conclusion, not the methodology. A VP doesn't care about your SQL query — they care about whether revenue is growing and what to do about it. Lead with the insight, show one clean visual to support it, and save the technical details for the appendix."

Ready to put it all together? Here's the two-week prep plan that covers every interview dimension.

Two weeks of focused preparation is all that separates "not quite ready" from "confident and composed." Here's the exact plan.

Prepared by Careery Team

Editorial Policy →

Reviewed byBogdan Serebryakov

Researching Job Market & Building AI Tools for careerists · since December 2020