← Back to Directory

Google interview notes, page 73 of 215

FAANG / Big Tech

Most of these Google write-ups come from Software Engineers (1,167), then Product Managers (686) and Data Scientists (125). The most reported rounds are the product sense round (667), the technical phone screen (660), the behavioral / leadership round (252) and the onsite coding round (214).

2146 interview notes · updated Oct 2026

Google·Data Scientist·Technical Phone Screen

May 2026
A technical phone screen for a Data Scientist role at Google, focused on a machine learning classification problem. The question was well-scoped but had some depth to it once you got past the surface-level setup.
  • Given a labeled dataset of 1,000 URLs marked as good or bad (dead links), walk through how you'd build and evaluate a classifier. Why might AUROC be a better evaluation metric than accuracy when the classes are imbalanced?

“I started with logistic regression which felt right, and the interviewer seemed fine with that.”

View Post

Google·Software Engineer·Technical Phone Screen

May 2026
Screened for a software engineer role at Google in Europe and came out with mixed results. The coding round was described by the recruiter as 'sufficient but average,' mostly because I needed a handful of hints to fix bugs even though my algorithm was solid. Behavioral went well. Now prepping for onsites with a month to fix my implementation sloppiness.
  • Implement an algorithm to solve a coding problem, arriving at the optimal time complexity solution.
  • Debug a method that could throw a NullPointerException and explain why it happens.
  • Identify why a piece of code would fail to compile.

“Got there on my own without hints for the actual approach, which felt good.”

View Post

Google·Software Engineer·Technical Phone Screen

May 2026
Google SWE interview with a debugging focus, specifically around build and CI systems. Pretty niche topic that I wasn't fully prepped for.
  • Given a broken or slow build/CI pipeline, how would you debug and fix it?

“I fumbled the opening a bit because I jumped straight into tooling instead of asking clarifying questions about what 'slow' actually meant.”

View Post

Google·Product Manager·Onsite - Product Sense / Strategy

May 2026
PM interview at Google with a single product analytics question about handling low survey response rates. Pretty short session from what I can tell, but the question itself had some depth to it.
  • You're running a user survey and you can't get enough respondents. What do you do?

“I went straight to incentives and sample size math, which felt too surface-level in retrospect.”

View Post

Google·Product Manager·Onsite - Product Sense / Strategy

May 2026
Google PM onsite focused entirely on product strategy, two big open-ended questions with follow-ups that kept coming. The kind of interview where you think you're done and then they hit you with five more angles.
  • How would you improve Google Maps?
  • What is the first Google Maps improvement you would cut if resources were constrained?
  • How would you design around driver distraction when adding new features to Google Maps?
  • How should Android compete more effectively against iPhone?
  • Which Android customer segment should Google prioritize and why?
  • How would you influence OEM partners to adopt consistent Android standards without taking away their flexibility?
  • What single metric would you use to measure whether Android is actually gaining ground against iPhone?

“I went broad too fast and had to walk myself back.”

View Post

Google·Product Manager·Onsite - Product Sense / Strategy

May 2026
Google PM interview, just the one question but it's a big one. The kind of thing that sounds fun until you're actually sitting there trying to not sound like you're pitching a sci-fi movie.
  • If you had unlimited resources at Google, what would you build and why?

“I rambled for probably two minutes before landing on anything coherent.”

View Post

Google·Software Engineer·Onsite - System Design / Architecture

May 2026
Google system design round focused entirely on building an L4 load balancer from scratch. Pretty deep dive, covered everything from data structures to failover. Left feeling okay about it but not great.
  • Design an L4 (transport layer) load balancer that handles around 1k QPS at normal load and up to 5k QPS at peak, where each backend server can handle roughly 10 QPS.
  • Walk through the different load balancing selection algorithms and their trade-offs, including round-robin, modular hashing, least-connections, and consistent hashing.
  • How would you handle backend health monitoring and dynamic membership changes in your load balancer design?
  • How would you estimate the capacity requirements, including server count, connection table size, and bandwidth needs?
  • How would you design for high availability and observability in this load balancer?

“The QPS numbers are what made this feel real.”

View Post

Google·Product Manager·Onsite - Behavioral / Leadership

May 2026
Went through a behavioral round at Google and got hit with a stakeholder expectations question that I thought I had a decent answer for. Spoiler: I did not.
  • Tell me about a time you delivered results that differed from what your stakeholders expected.

“I went in thinking this was a classic 'managed up' story and picked an example where I technically hit the goal but the scope had shifted.”

View Post

Google·Software Engineer·Technical Phone Screen

May 2026
Google SWE interview with a market matching engine problem that looked like a coding exercise but turned into a pretty involved simulation with aggregation logic. The FizzBuzz framing is cute until you realize how many edge cases are buried in there.
  • Build a simulated matching engine where 'Fizz' orders are sells and 'Buzz' orders are buys. Orders come in FIFO priority. A match occurs when a buyer's price is at least the seller's ask. Trade price is the midpoint of the two. Process each new order greedily against the earliest compatible resting orders. Partial fills leave residual quantity. Aggregate all trades between the same buyer-seller pair into a single record with summed quantity and volume-weighted average price.

“I spent way too long on the matching loop and almost forgot about the aggregation step entirely.”

View Post

Google·Software Engineer·Technical Phone Screen

May 2026
Google SWE coding round, one problem the whole time. Array manipulation with a follow-up that I didn't fully see coming. Not sure how I did.
  • You're given an integer array where each element represents the weight of a chain link. Remove exactly one link, then check if the remaining chain can be split into two contiguous segments of equal total weight. Return true if any such removal exists. Follow-up: return all valid (removed index, cut position) pairs.

“The base problem clicked pretty fast.”

View Post