INTERVIEW GUIDE

Google Data Analyst Interview: Questions & Process

Google's Data Analyst interview leans on three things: writing clean SQL under time pressure, reasoning soundly about statistics and experiments, and turning vague product questions into measurable metrics. Expect a recruiter screen, a technical phone screen, and a 4-5 round virtual onsite that also tests 'Googleyness.'

The interview process

1. Recruiter screen — ~30 min call

Tests: background, role fit, comfort with SQL and stats, logistics and timeline

2. Technical phone screen — 45-60 min, shared editor

Tests: live SQL against a sample schema plus a few applied statistics questions

3. SQL / analytics deep-dive — 45 min onsite

Tests: joins, window functions, and aggregation on a realistic dataset; explaining your query choices

4. Product / metrics case — 45 min

Tests: defining success metrics, designing an A/B test, and diagnosing why a metric moved

5. Behavioral / Googleyness — 45 min

Tests: collaboration, handling ambiguity, intellectual humility, and the impact of your work

Questions you're likely to get

Technical

Given a table of search sessions, write SQL to find the daily count of users who searched more than 3 times.
Use a window function to rank queries by volume within each country.
Write a query to compute 7-day retention for users who signed up last month.
Explain the difference between a p-value and a confidence interval to a non-technical PM.
How would you detect and handle outliers in a metric before you report it?

Role-specific

How would you measure the success of Google Maps' 'save your parking location' feature?
Daily active users on YouTube dropped 4% week over week — how do you investigate?
Design an A/B test for a new Search results layout. What's your primary metric and what are your guardrails?
What metrics would you track for Google Photos, and which single one matters most?
How do you decide whether a 2% lift in click-through is real or just noise?

Behavioral

Tell me about a time your analysis changed a decision.
Describe a time you had to explain a complex finding to a non-technical audience.
Tell me about a time you were handed an ambiguous question with no obvious data to answer it.
Describe a time you disagreed with a stakeholder about what the data actually showed.

How to answer (worked examples)

Daily active users on YouTube dropped 4% week over week — how do you investigate?

Resist guessing. First clarify the metric definition and time window, then rule out data and logging issues. Segment the drop (geography, platform, new vs. returning, app version) to localize it, check for seasonality or a known release, then form one hypothesis and say exactly what query or data you'd pull to confirm it. Google rewards a structured, MECE breakdown far more than a fast guess.

Design an A/B test for a new Search results layout.

State the hypothesis, then the unit of randomization (the user). Choose a primary metric tied to the goal — successful sessions, not just clicks — plus guardrail metrics like latency and ads revenue. Mention sample size and power, how long you'd run it to avoid peeking, and call out novelty effects. Showing you protect against false positives is the signal.

Tell me about a time your analysis changed a decision.

STAR. Situation: the decision in play and the stakes. Task: the question you owned. Action: the specific analysis and how you pressure-tested it. Result: the decision that changed and the measurable impact — and one sentence on what you'd validate further next time.

What Google looks for

Fluent SQL — joins, window functions, and aggregation without hand-holding
Structured problem-solving — you break an ambiguous question into a clear framework
Statistical rigor — you reason correctly about significance, sample size, and bias
Product sense — you tie metrics to user and business goals, not just numbers
Clear communication — you can explain a finding simply to non-technical partners
Red flag: jumping to a conclusion without checking data quality or stating assumptions

FAQ

Does Google ask SQL in the Data Analyst interview?

Yes — live SQL is a core component. Expect to write joins, aggregations, and window functions against a sample schema and to explain your choices, not just produce a correct answer.

Do I need to know statistics?

Yes. A working command of A/B testing, statistical significance, sampling, and confidence intervals comes up in both the technical screen and the product case.

What is 'Googleyness'?

It's Google's culture-fit signal: comfort with ambiguity, collaboration, intellectual humility, and bias toward impact. A dedicated behavioral round probes for it.

How long does the process take?

Usually 4-8 weeks from recruiter screen to offer, with the hiring-committee review adding time after the onsite.

Do they ask Python or R?

SQL and analytics are the core. Some roles include light scripting, but you won't be expected to grind algorithmic coding the way a software engineer would.

Practice the Google analytics case out loud

You can read every SQL pattern and still freeze when an interviewer asks why DAU dropped. Rehearse this exact loop — the metrics case, the A/B test design, the Googleyness round — with OfferLoop's realtime voice coach before the real thing.

Practice this interview out loud →

All Google interviews →Google Software Engineer interview →Google Data Scientist interview →Google Data Engineer interview →Google Machine Learning Engineer interview →

OfferLoop is an independent interview-practice tool and is not affiliated with, endorsed by, or sponsored by Google. All company names and trademarks are the property of their respective owners.

Interview formats vary by team, level and year, and this guide is compiled from general knowledge of publicly discussed hiring processes — treat it as preparation material, not an official description of Google's current process.