3 ms·

Had a similar experience. The recruiting part of the process went pretty smoothly, until I got to the onsite. Had 5 on-site interviews (excluding lunch), all ex

by typon 4y ago

Had a similar experience. The recruiting part of the process went pretty smoothly, until I got to the onsite. Had 5 on-site interviews (excluding lunch), all exactly the same format: walk in the room where my interviewer is waiting with a question written on the white board, they say "hello" to me and I'm expected to read the question, clarify it a little bit and start coding. Not a single interviewer cared about my past experience or what kind of a person I am. I guess they're relying on the lunch interview for that signal? Anyway I ended up getting rejected and the feedback was that I need to practice my algorithms more. (All the questions were slight variations on popular leetcode questions)

skeeter2020 4y ago
>> and the feedback was that I need to practice my algorithms more. I'm optimistic I will live long enough to see the death of this type of interview. Great gate-keeper question, easy to mark, next to zero signalling value as to ability to problem solve or program. Sigh...
- lostdog 4y ago
  What would you consider to be a good interview question?
Tyr42 4y ago
(Googler, my opinions are my own, etc) When interviewing we are instructed to write up the interview question and answer and score on a rubric with citations into the interview notes. These notes then go to the hiring committee who makes the decision without knowing the human side at all. I think the point is something like this https://www.theguardian.com/women-in-leadership/2013/oct/14/blind-auditions-orchestras-gender-bias https://www.theguardian.com/women-in-leadership/2013/oct/14/... where the hiring committee is intentionally blinded. There just isn't anywhere in the notes to record "what kind of a person you are". There is a separate interview which tries to capture that, but the rest of this is in the name of avoiding bias, and maybe being more lenient with someone who went to the same school as you, or makes you laugh or whatever.
- typon 4y ago
  that reasoning is understandable actually. Given Google's success, maybe this is the right way of approaching the hiring process over the long term at the scale of tens of thousands of employees. However, that kind of culture is just not for me. I ended up getting hired at a small startup working on cutting edge technology and have really enjoyed my time here. So in a way, it turned out well for both me and Google.
bena 4y ago
I've done a live on-site and a virtual on-site during COVID. I was mostly stationary during the live on-site. I was taken to a room and the interviewers came to me. Some chit-chat, then given the problem, then working on the problem. A break for lunch, then I was taken to a room near where the first was for the rest of my interviews. Much the same deal. I stayed in the room while the interviewers cycled through. I was told that the lunch portion wasn't part of the interview and the person escorting me around didn't have any influence on the hiring decision. That may or may not be true, but it is what they told me. About a week after, I was informed of my rejection. I'm fairly sure I know one interview I particularly bombed. The question involved something I had done before, but a while ago and I floundered a bit trying to recall all the bits and bobs of the syntax and structure. The other four I felt could go either way. During the virtual on-site, it was just being passed off to various Google Meet rooms. I think I wrote code in a shared Google Doc. I felt ok in most of them. None particularly stuck out. A few days after the interview, I was informed I was being passed to the hiring committee. And then a few days after that, I was informed the hiring committee passed on me. All the questions I got were vaguely LeetCode-ish. They were all of the template: Here's a vague problem with loose requirements. I'd ask some questions about particulars, "Can I assume this?", "Is this true for all?", "Is this sorted?", "Can I go to the bathroom?" "No, now I really have to go, last time was to avoid answering a question" etc. Then I'd write up a solution in my language of choice, trying to explain why I was doing what I was going while doing it. Just stream of conscious style stuff. Occasionally, I'd get comments or questions as I talked. Stuff about assumptions I was making or why I was trying this method. Then if there was time and they were satisfied with my solution, they'd ask me to essentially scale it up. "Ok, that works when you can hold the dataset in memory, but what if the dataset was three bazillion petabytes?" I personally enjoy puzzles and challenges, so I found the process fun even though the fact that it is an interview makes it a little more. But since I have also had a job both times I interviewed, the stakes are a bit lower for me, which also helps.

by typon 4y ago

skeeter2020 4y ago

>> and the feedback was that I need to practice my algorithms more. I'm optimistic I will live long enough to see the death of this type of interview. Great gate-keeper question, easy to mark, next to zero signalling value as to ability to problem solve or program. Sigh...

lostdog 4y ago
What would you consider to be a good interview question?

Tyr42 4y ago

(Googler, my opinions are my own, etc) When interviewing we are instructed to write up the interview question and answer and score on a rubric with citations into the interview notes. These notes then go to the hiring committee who makes the decision without knowing the human side at all. I think the point is something like this https://www.theguardian.com/women-in-leadership/2013/oct/14/blind-auditions-orchestras-gender-bias https://www.theguardian.com/women-in-leadership/2013/oct/14/... where the hiring committee is intentionally blinded. There just isn't anywhere in the notes to record "what kind of a person you are". There is a separate interview which tries to capture that, but the rest of this is in the name of avoiding bias, and maybe being more lenient with someone who went to the same school as you, or makes you laugh or whatever.

typon 4y ago
that reasoning is understandable actually. Given Google's success, maybe this is the right way of approaching the hiring process over the long term at the scale of tens of thousands of employees. However, that kind of culture is just not for me. I ended up getting hired at a small startup working on cutting edge technology and have really enjoyed my time here. So in a way, it turned out well for both me and Google.

bena 4y ago

I've done a live on-site and a virtual on-site during COVID. I was mostly stationary during the live on-site. I was taken to a room and the interviewers came to me. Some chit-chat, then given the problem, then working on the problem. A break for lunch, then I was taken to a room near where the first was for the rest of my interviews. Much the same deal. I stayed in the room while the interviewers cycled through. I was told that the lunch portion wasn't part of the interview and the person escorting me around didn't have any influence on the hiring decision. That may or may not be true, but it is what they told me. About a week after, I was informed of my rejection. I'm fairly sure I know one interview I particularly bombed. The question involved something I had done before, but a while ago and I floundered a bit trying to recall all the bits and bobs of the syntax and structure. The other four I felt could go either way. During the virtual on-site, it was just being passed off to various Google Meet rooms. I think I wrote code in a shared Google Doc. I felt ok in most of them. None particularly stuck out. A few days after the interview, I was informed I was being passed to the hiring committee. And then a few days after that, I was informed the hiring committee passed on me. All the questions I got were vaguely LeetCode-ish. They were all of the template: Here's a vague problem with loose requirements. I'd ask some questions about particulars, "Can I assume this?", "Is this true for all?", "Is this sorted?", "Can I go to the bathroom?" "No, now I really have to go, last time was to avoid answering a question" etc. Then I'd write up a solution in my language of choice, trying to explain why I was doing what I was going while doing it. Just stream of conscious style stuff. Occasionally, I'd get comments or questions as I talked. Stuff about assumptions I was making or why I was trying this method. Then if there was time and they were satisfied with my solution, they'd ask me to essentially scale it up. "Ok, that works when you can hold the dataset in memory, but what if the dataset was three bazillion petabytes?" I personally enjoy puzzles and challenges, so I found the process fun even though the fact that it is an interview makes it a little more. But since I have also had a job both times I interviewed, the stakes are a bit lower for me, which also helps.