Google Senior+ Data Scientist Interview Experience — Rejected Everywhere Else Before Landing an All-Strong-Hire DSR Offer

I've read some experience posts on the forum before, and this is my way of giving back — my first post, sharing some of my own experience here. Some background first: I'm at a big company, I've been coasting for several years, and I felt like I was wasting away, so I planned to switch jobs earlier this year. It didn't go well though — I got rejected at the hiring manager round, then coding/SQL rounds, then onsite, then the final round, before finally landing an offer at Google. So if you're also feeling lost starting your job search like I was, this post might help a bit.

I wasn't short on interviews, but the whole DS job market was down — too many candidates, not enough openings — so the interviews didn't go smoothly at all. Let me start with the Google timeline and roughly what the interview covered:

Timeline:

• End of August: submitted my resume
• Early October: HR called me directly, while I was still at work... Two teams had picked me, and after some internal discussion they had me interview for DSR. I felt like I hadn't done enough homework here — I'd just applied to a few roles randomly. I didn't even know Google's internal DS org had different tracks.
• Late October: phone screen, two rounds, heard back a week later that I passed.
• Mid-November: virtual onsite (VO), three rounds
• Two days after the onsite I got the news I passed, multiple teams interested
• Late November: team match

I'd interviewed at Google once before, five years ago, and got knocked out at the onsite. As for how this round felt — maybe because of my years of experience now, they didn't ask any basic probability or distribution derivations anymore. It was all business use-case application. At most they asked something like how to run an AB test and how to reduce variance; everything else was applied questions about measuring performance — things like experiment design (e.g. how to test latency's impact on some metric), causal inference (can you explain how you'd use propensity score matching in this scenario?), and on the modeling side, mixed models, random/fixed effects, and so on. On the product side there were the usual analytics questions — how to diagnose a change in a ratio metric, how to define a metric for the downstream impact of YouTube latency, that kind of thing. The coding round was honestly insulting to my intelligence — way too easy, to the point that I had 15 minutes left with nothing to do and just made small talk. It covered things like simulating a uniform distribution, aggregating a table to achieve some goal, and some operations on the Fibonacci sequence.

My overall impression is that the questions were pretty standard, and honestly easier than the ball-drawing probability questions they gave me a few years back. I can't go into more detail on the specific questions, and since I don't grind practice problems and also can't see most of the experience posts, let me offer some unconventional prep advice instead: read books, and ask ChatGPT whenever you don't understand something.

Everyone probably has their own way of prepping for coding, and for DS it's basically just SQL + Python, generally not that hard, so I won't get into it here.

All told I read at least four or five books. Cramming obviously doesn't work, so this definitely wasn't a last-minute strategy — I read these in my spare time because they're useful for the job, and it turns out they're also very useful for interviews. I'd recommend them to everyone; at least for the DS track, they'll definitely get you through.

I'd strongly recommend George Casella and Roger L. Berger's "Statistical Inference." It goes from basics to advanced, and it lays out all the foundations and principles of statistics really clearly. There were parts of this textbook I didn't understand back in school, but going over it again and again, I understand a lot more now, and it's also really helpful for reading papers. A lot of the concepts it covers — bootstrap, delta methods — come up a lot in interviews. For example, when I got asked "can bootstrap help reduce variance," the book has detailed examples and comparisons on that, which was very useful.

Causal inference: I've personally read a lot of books and papers on this topic. If you're just prepping for interviews, Scott Cunningham's "Causal Inference: The Mixtape" is enough — it has tons of examples and walks through a lot of great papers. If you're into causal inference research more broadly, there's a Stanford lecture series on YouTube called "Machine Learning & Causal Inference: A Short Course" — it's an absolute treasure trove, I really love it.

Product analytics: This is what I'm weakest at. I deal with it every day at work, but honestly, interview-wise it's really just about knowing the playbook. I personally really don't like Meta's interview style — I interviewed at several companies, and any time I ran into someone who came from Meta, I bombed it. I genuinely couldn't find the pattern — sometimes I got the metrics wrong, sometimes the HR feedback was that I didn't clarify the question enough. After talking to a friend who works at Meta, I suddenly realized I'd been acting like a total idiot before that. I'd recommend the following two books — they're not just useful for interviews, they're useful for the job too:

Kevin Huo and Nick Singh's "Ace the Data Science Interview" — the product sense section is written so well. It lays out clearly how to approach each type of question and how to analyze it, and it's a framework you can genuinely apply directly. Tons of examples.

"Lean Analytics" by Alistair Croll and others: fundamental product analytics. Different industries have their own metrics systems — what matters most for a two-sided marketplace, what SaaS cares about, how to analyze a social network, how to define key metrics, and so on. No matter how much AI changes things these past few years, the underlying business models have stayed the same for over a decade, so this is a way to handle any variation with one solid foundation.

AB testing: last one to mention — one book is enough here, "Trustworthy Online Controlled Experiments" by Ron Kohavi. It's a pretty famous book, and every time I reread it after a while I get something different out of it. Ron Kohavi is really good at explaining confusing concepts, and he's written a lot of papers too, which were also very worthwhile — if you're interested, it's worth searching them out.

On ChatGPT: I'm genuinely thankful for the progress of technology — whatever I don't understand, I just ask, and it can summarize the question and the reasoning very accurately. I usually use it as a first step to carefully research a specific company and the business chain I'll be interviewing for — things like revenue sources, customer sources, industry characteristics, and so on. Combined with the industry-wise metrics from the books above, that basically gets you the main product metrics. AB testing and product analytics, which are closely tied to metrics, stop being a problem after that. Because I'd read so much, I could naturally anticipate the product-related questions and the traps an interviewer might dig, and after discussing them with ChatGPT, there generally weren't any major issues left.

Last thing — my personal takeaway: the whole job search process made me grow faster than my last several years of actual work — maybe that just says how watered-down my org is... But throughout the whole prep process, I could clearly feel my learning ability improving and the concrete payoff from adjusting my strategy. To sum it up:

If you have the connections, talk to friends at your target companies as much as possible and see how they approach things. Some blogs online are really well written — I'd recommend Meta's; an HR rep gave me a prep doc for an earlier interview, and looking back now it's actually a goldmine.

After every failure, make sure — really make sure — you figure out why you failed. Most of the time you actually already know right after the interview where you did badly. If you happen to get a kind HR rep, seize the chance to ask for feedback. Make sure you never fall into the same trap twice.

Personally I don't recommend endlessly grinding through experience posts — even if you get lucky and hit the exact same question, you might not actually understand it. If possible, seeing through the question to the underlying thing it's testing and the strategy for solving it is the real answer.

Interviewing takes skill — there are playbooks and methods you can follow, but all of that only comes through practice. Unless you're some kind of genius who passes every single interview, you're bound to miss out on some companies, so if you get the chance, interview as much as you can and keep a learning mindset. I went from being completely lost at the start to making it to the final round and getting all-strong-hire feedback, and I got there through countless mistakes along the way. Without all those earlier missteps, there's no way I could have landed this with just my own ability — but I'd rather the people who come after me learn from what I went through and take fewer detours.