Google Interview Questions — Page 205

Google·Product Manager·Onsite - Product Sense / Strategy

Jul 2026

Got a Google Maps PM question that sounds deceptively simple but ends up being this sprawling business justification exercise. The setup is weirdly adversarial and I kind of liked it.

You're a PM on Google Maps and your manager walks in asking you to shut it down with no explanation. How do you build the case for why Google should keep investing in Maps?

“My first instinct was to go straight to revenue and I think that was the wrong entry point.”

View Post

Google·Software Engineer·Technical Phone Screen

Jul 2026

Google data engineer screen, just one technical question about file formats. Pretty short session, felt more like a warm-up than a full round.

What are the key differences between Parquet and Avro as data storage formats?

“I knew the basics but fumbled a bit on when you'd actually pick one over the other in practice.”

View Post

Google·Software Engineer·Onsite - Coding / Algorithms

Jul 2026

After completing coding rounds for a Software Engineer role at Google, the recruiter came back saying the signals were strong for a mid-level position but thinner for the level above, and offered an extra coding round to give a clearer read. Now sitting in limbo wondering what the extra round looks like, what it actually changes, and whether tanking it means starting over or just sliding down a level.

If you're offered an additional coding round to better assess for a higher level, what should you expect it to look like compared to previous rounds?
If the extra round goes well, what are the realistic odds of being cleared for the higher level versus the lower one?
If you underperform in the additional round, are you still considered for the lower level or does a poor result knock you out entirely?

“Genuinely unsure about this.”

View Post

Google·Data Scientist·Technical Phone Screen

Jul 2026

Deep technical screen for a DS role at Google, basically one massive ML system design question broken into five parts. The whole thing felt like a take-home prompt read aloud, and they expected you to just... go. No warm-up, no softball.

Given minute-level battery telemetry for a specific device currently at 37% charge, how would you implement a baseline time-to-empty estimator using piecewise linear interpolation on the discharge curve? How do you handle irregular timestamps, short charging periods, and screen-off segments?
Name at least 8 engineered features you would add to improve the time-to-empty prediction beyond the raw discharge curve, and explain the expected direction of each feature's effect on the estimate.
For a device model with no historical data (cold start), how would you transfer discharge curve priors from other devices? Define your similarity metric, how you combine neighbor curves, and how you quantify prediction uncertainty.
How would you design an offline evaluation protocol for this predictor? What metrics would you use as primary and secondary, and how would you guard against the model systematically underestimating time-to-empty?
After the model is live, how would you monitor for distribution shift in the discharge patterns, and what would an automatic recalibration procedure look like?

“This is where I spent way too long.”

View Post

Google·Data Scientist·Technical Phone Screen

Jul 2026

Google data scientist interview that went deep into statistical sampling theory, way deeper than I expected for a DS role. The whole session was basically one long question about truncated normals and I was not fully prepared for the numerical precision angle.

Define the truncated normal distribution for Z given a < Z < b where Z is standard normal. Write out the normalized pdf and cdf.
Design efficient samplers for three specific truncated normal cases: (i) lower tail cutoff at 1 with no upper bound, (ii) a very narrow interval from 4 to 4.05, and (iii) a lower tail cutoff at 4 with no upper bound. For each case, compare inverse-CDF sampling against naive rejection sampling from a standard normal.
Propose a sampling algorithm suited for extreme tail regions, such as an exponential or half-normal proposal with acceptance-rejection. Walk through the construction and why it works better than the alternatives.
How do you handle numerical stability when evaluating the normal CDF and its inverse near machine precision, especially in extreme tail regions?
How would you validate that your truncated normal sampler is actually correct? Describe a testing approach.

“This part was fine, I knew the forms.”

View Post

Google·Data Scientist·Technical Phone Screen

Jul 2026

Google DS interview with a meaty growth problem centered on Workspace Chat. The whole thing was essentially one long case question about driving MAU growth, and they wanted you to go pretty deep on the experimental side of things.

Google Workspace Chat has low adoption. How would you build a plan to grow monthly active users, including which metrics to track, what hypotheses you'd form, how you'd structure an A/B testing roadmap, and what success looks like?

“This one sprawled in a way I wasn't ready for.”

View Post

Google·Software Engineer·Technical Phone Screen

Jul 2026

Got a machine learning flavored coding question at Google for a software engineer role. The problem was about building a bigram language model from scratch, which sounds straightforward until you get to the follow-up about scaling it.

Given a set of token sequences as training data, build a next-word prediction model that counts how often each word is followed by another word, then at inference time sample the next word proportionally to those observed frequencies.
How would you handle this system if the vocabulary is very large or each word has a huge number of distinct successors, given memory and latency constraints?

“The bigram part was fine, just a frequency map.”

View Post

Google·Product Manager·Onsite - Multi Round

Jul 2026

Google PM loop with a mix of estimation, business strategy, and a full product walkthrough. The questions ranged from quick fermi-style math to a pretty deep case on Netflix's global business model, which felt like a lot to cover in one session.

Estimate the number of views a YouTube video about cats would get in the first week after being uploaded.
Estimate Netflix's total annual revenue.
Compare what drives revenue for Netflix in the U.S. versus internationally, and explain how those differences should shape Netflix's strategy going forward.
Walk me through a product project you led end to end, covering the problem, your hypothesis, user research, how you defined success, roadmap decisions, the launch, and what you learned afterward.

“I anchored on total daily YouTube views, tried to carve out a slice for animal content, then adjusted for recency since new videos don't immediately surface.”

View Post

Google·Software Engineer·Onsite - Behavioral / Leadership

Jul 2026

Behavioral round at Google for a software engineering role. Five questions, all with aggressive follow-ups demanding numbers, root causes, and proof. The kind of interview where vague answers visibly die on the table.

What was the biggest challenge you faced, and how did you measure the outcome?
Describe a difficulty you overcame. How did you find the root cause, and what did you do afterward to prevent it from happening again?
What have you done in the past year to improve yourself professionally?
Tell me about a time you disagreed with someone or had a conflict at work. How did you get them on board with your view?
What's a project you're most proud of, and why?

“The follow-up is where this gets uncomfortable.”

View Post

Google·Software Engineer·Onsite - System Design / Architecture

Jul 2026

System design round at Google for a software engineer role. The prompt was about building a school messaging platform where staff send notifications to guardians who have to confirm they saw it. Solid problem with a lot of moving parts.

Design a system that lets school staff send messages to students' guardians, where guardians must acknowledge receipt. Cover sender and recipient management, reliable delivery, acknowledgement tracking, reminders, escalation, audit trails, privacy and consent, and reporting.

“This one sprawled fast.”

View Post