That might happen to lesser companies, but not these companies, who are crushing it because they're so full of awesome talent because they only hire the best that hiring methods that wouldn't work for other companies will work for them.
yeah, this article is basically stuff that actually happens that seems absurd even while you're going through it. I love the multiple rooms one, happens all the time!
It should have had two more:
Interview the person while they're eating lunch, but tell them "it doesn't matter, unless you really do something bad."
Or the converse, put them in an interview from 9-5, but don't offer them lunch "because this one guy is only available at noon."
I didn't even read the article because I know there is no secret.
We don't hire the best people. We hire average people, some good people, some great people, some mediocre people, and some people who aren't very good at their jobs. Some of those mediocre and average people turn out to be great people after all. There's no secret to it. We like telling ourselves there is a secret because everyone likes to feel special and we think if we can mirror other companies that we think are successful then we can be successful just like they were.
In the end, you have people working with people. If you're fortunate, you get the right people interviewing the right person at the right time, and you get a great hire. At best, you can filter out the obviously bad ones. The rest, you'll find out about as you go. Some will surprise you, some will meet your expectations, and some will be disasters. Fire the bad ones; coach and mentor the average ones; keep track of the great ones.
Go build some shit with people you like building shit with and stop trying to figure out how to make hiring objective. You can't.
I'd like to believe that both paxunix and andrewsteward wrote those comments after having read the article, but pretended like they didn't for the Internet points. Actually makes the comments a bit funnier.
> Go build some shit with people you like building shit with and stop trying to figure out how to make hiring objective. You can't.
Yes you can. It's quite easy, make people build something small to get hired, that eliminates 99% of submissions who are just lying about their skills and posing and with code in hand, it's quite easy to decide who's full of shit and who isn't. Resume's are useless, demand code, real programmers program.
This is a good summary
"Go build some shit with people you like building shit with and stop trying to figure out how to make hiring objective. You can't."
So true. I had an Amazon interview where I wanted to hang up. There were moments where I would get questioned or berated for not using such great AWS tech like Opsworks and Cloudformation. Instead of being impressed with my completely valid approach they fixated on the fact I didn't use their stuff and it became a bit combative/snarky. So Amazon certainly isn't for me. Thanks for EC2 and two day shipping though.
I had a phone screen from Google I wanted to hang up on because I knew where it was going. They reached out to me for an SRE role in Australia. After a few unscheduled calls at poor times (at the DC doing work) we setup an "interview". It was just a screen where they ask you to rate your self on a bunch of random technologies; between never heard of it and wrote the book. Rated myself 4-6 on most haha. The " interviewer" seemed completely thrown off by this. Expected not to hear back and didn't.
I never understood the rate yourself bit. Is it designed to fuck with your head? "So you think you are good? We'll show you. How about this obscure trick question from C++17 spec -- see not as good as you thought"
Meh, I ask it on a couple technologies now and again but only for broad tech. It does two things for me. One is that for some of my positions, I need a candidate to be good at some but not all things my team does, like windows, linux or net ops. So it helps me focus my questions on what the candidate feels comfortable with. It's not a trick.
The other thing that you can often see with it is how much a person doesn't know that they don't know. I don't often get someone telling me that they are a 9/10 in linux but generally when they do, they can't set an ip from the CLI, or use top. I do want to see people who understand what they know and what they don't. I realize the question is pretty arbitrary so it's one data point in a set but it can be useful to understand how someone sees their own competency.
> I do want to see people who understand what they know and what they don't
Ok that's fair, I can see that point.
I usually just probe by asking details about the projects they've done etc.
Also found there is sometimes an inverse corellation between those who rate themselves high and their abilities and vice verse rate themselves low and their abilities (there is a name for this phenomenon, but I don't recall it). The idea is those who really know a lot, also know how much they don't know.
If someone asked me "how well you know the Linux kernel", I would say not at all. Because I had to read the source to some parts of it, had to tune it for realtime charactestics and have seen the depth of the rabbit hole so to speak, and realized I don't know anything compared to how much there is to know about it.
Yes so the ranking is arbitrary but I know that you've made it to the top of the rabbit hole and looked down it once. Some people are at the edge of the forest and don't even know there is a rabbit hole. That's what the question is for.
> I usually just probe by asking details about the projects they've done etc.
I do this too, but that tells you if the person has a matching view of their abilities with their actual skills.
When you ask them to configure something, do you let them sit in front of a laptop with man pages, google and stack exchange?
Or do you expect them to memorize all your most used command line tools?
Do you expect them to remember the parameters to ifconfig, or do you ask them to craft a script/command line to print only the MTU for each interface?
When you ask them about top, do you expect them to remember what top is? or do you allow them to google for top, and ask them to craft a command line (or shell script) that finds which user is using the most memory on the system?
Genesis10, recruiting on behalf of Bloomberg, did that shit to me several years ago. The interviewer decided to gauge my knowledge of Unix/Linux and POSIX programming by essentially quizzing me on POSIX manpages over the phone.
I don't really ask people to configure things. I ask if you understand how DHCP works, and how TCP works, how top works (ie info is in proc . . . I volunteer laptop at the start of interviews but most people never use it. I try to stray away from asking for a list of facts.
Often I will ask a question about making a big change and I want to know how they go about it, is there error checking, is the a list of failures that can be worked later, if it's big scale are things run in parallel, are parallel processes run in a way that would max out the boxes resources? I don't need all of those because everyone answers those questions differently but I want to know how you think and assume you can look up arguments later. Or that I can teach you how to use a man page depending on the level of hire.
A huge majority of our stack is proprietary so you won't know how to use it anyway. So if you know how to think and you know how to learn then you should be fine.
(Tedious disclaimer: my opinion only, not speaking for anybody else. I'm an SRE at Google.)
I can't tell you about other companies, but ours is pretty straightforward: you tell us what your strongest areas are, we'll give you interviews that line up with those areas. There's no strict guarantees and you should expect a spread of different questions, but whatever it is that you're good at, expect us to offer you a chance to show us that you're good at it.
It's probably only a useful technique at companies which are big enough and flexible enough to find a use for you, whatever your skill set happens to be. For a company that only has one job opening and needs a specific sort of person to fill it, this sort of pick-your-interview-subject approach doesn't help them.
(Trick questions are useless: they primarily serve to make the interviewer feel clever, not to demonstrate the abilities of the candidate)
Ok makes sense. I should also apologize, Google didn't really give me trick questions when I talked to them 4 years ago or so (but other companies did), so didn't mean to imply they did as well.
Google had a list of technoliges as you described, and I was asked to rate myself on each one. I think it would have been slightly better if they explain why or how it is used.
> I can't tell you about other companies, but ours is pretty straightforward: you tell us what your strongest areas are, we'll give you interviews that line up with those areas. There's no strict guarantees and you should expect a spread of different questions, but whatever it is that you're good at, expect us to offer you a chance to show us that you're good at it.
My one experience with getting to a Google onsite is that, despite repeatedly telling various recruiters my strongest areas, they ended up interviewing me for what they wanted me to do rather than what I was good at. In fact, one of my interviewers literally had to throw out all his prearranged questions and make shit up as he went because my background and experience were so misaligned with what he had been added to my loop to ask me about.
Having a lot of confidence in your abilities is arguably a pretty important quality in an employee. I've encountered a lot of engineers who were way less effective than they could have been, mainly because they were too timid to speak up, to take initiative, to lead.
So maybe they were right to reject someone who answered 4-6 in most categories. Either that person isn't a good fit skill-wise, or they are a good fit skill-wise, but not personality-wise.
(Tedious disclaimer: my opinion only, not speaking for anybody else. I'm an SRE at Google.)
You did not have a phone screen. What you are describing is the initial contact with a recruiter, to establish what roles you're interested in, explain how the process works, and arrange dates. Had you proceeded past this, you would probably have had a phone screen next, which would be with an engineer giving you a technical problem to solve.
I also rated myself 4-6 on most categories, and was hired; that's an entirely reasonable score on the scale they give. I don't know what happened in your conversation with the recruiter, but it sounds like you made it clear you weren't interested and that was the end of it.
You are probably correct. It's been a few years but I do recall not being super impressed with the questions after the hassal of a bit of phone tag and a scheduled call. He likely picked up on my disinterest. I could have sworn though that there was this ackwardness based on my responses. Like he couldn't reconcile my answers with my profile.
My not be considered a "phone screen" in the process but if it walks like a duck and quacks like a duck.. :)
Basically, this interview process allows you to weed out candidates who
1) disapprove open floor plan work environment, sitting next than six feet away from co-worker (yeah let's cramp twenty people together around so we can hear each other yelling at the headphone)
2) markers are either missing or used up and no one replaces them. oh you want to project your screen? good luck find a working remote or HDMI cable.
108 comments
[ 3.4 ms ] story [ 177 ms ] threadEvery single article on the site has titles that would give Buzzfeed run for its money.
http://thecooperreview.com/
It should have had two more:
Interview the person while they're eating lunch, but tell them "it doesn't matter, unless you really do something bad."
Or the converse, put them in an interview from 9-5, but don't offer them lunch "because this one guy is only available at noon."
Both have happened to me while interviewing. :)
We don't hire the best people. We hire average people, some good people, some great people, some mediocre people, and some people who aren't very good at their jobs. Some of those mediocre and average people turn out to be great people after all. There's no secret to it. We like telling ourselves there is a secret because everyone likes to feel special and we think if we can mirror other companies that we think are successful then we can be successful just like they were.
In the end, you have people working with people. If you're fortunate, you get the right people interviewing the right person at the right time, and you get a great hire. At best, you can filter out the obviously bad ones. The rest, you'll find out about as you go. Some will surprise you, some will meet your expectations, and some will be disasters. Fire the bad ones; coach and mentor the average ones; keep track of the great ones.
Go build some shit with people you like building shit with and stop trying to figure out how to make hiring objective. You can't.
You know:
"Why my startup failed" (cause I built something no-one wants)
"Why we switched from this technology to that technology" (because the old one sucked but the new one will solve all our problems)
"How we built a startup that fixed hiring" (they didn't)
"How to hire great people" (no-one knows)
"Why we were rejected from YC" (can an idea be that bad?)
"Our experience of interviewing at YC" (we were scared, it didn't take long, we got accepted!)
etc etc
Yes you can. It's quite easy, make people build something small to get hired, that eliminates 99% of submissions who are just lying about their skills and posing and with code in hand, it's quite easy to decide who's full of shit and who isn't. Resume's are useless, demand code, real programmers program.
The other thing that you can often see with it is how much a person doesn't know that they don't know. I don't often get someone telling me that they are a 9/10 in linux but generally when they do, they can't set an ip from the CLI, or use top. I do want to see people who understand what they know and what they don't. I realize the question is pretty arbitrary so it's one data point in a set but it can be useful to understand how someone sees their own competency.
Ok that's fair, I can see that point.
I usually just probe by asking details about the projects they've done etc.
Also found there is sometimes an inverse corellation between those who rate themselves high and their abilities and vice verse rate themselves low and their abilities (there is a name for this phenomenon, but I don't recall it). The idea is those who really know a lot, also know how much they don't know.
If someone asked me "how well you know the Linux kernel", I would say not at all. Because I had to read the source to some parts of it, had to tune it for realtime charactestics and have seen the depth of the rabbit hole so to speak, and realized I don't know anything compared to how much there is to know about it.
> I usually just probe by asking details about the projects they've done etc.
I do this too, but that tells you if the person has a matching view of their abilities with their actual skills.
Do you expect them to remember the parameters to ifconfig, or do you ask them to craft a script/command line to print only the MTU for each interface?
When you ask them about top, do you expect them to remember what top is? or do you allow them to google for top, and ask them to craft a command line (or shell script) that finds which user is using the most memory on the system?
Often I will ask a question about making a big change and I want to know how they go about it, is there error checking, is the a list of failures that can be worked later, if it's big scale are things run in parallel, are parallel processes run in a way that would max out the boxes resources? I don't need all of those because everyone answers those questions differently but I want to know how you think and assume you can look up arguments later. Or that I can teach you how to use a man page depending on the level of hire.
A huge majority of our stack is proprietary so you won't know how to use it anyway. So if you know how to think and you know how to learn then you should be fine.
I can't tell you about other companies, but ours is pretty straightforward: you tell us what your strongest areas are, we'll give you interviews that line up with those areas. There's no strict guarantees and you should expect a spread of different questions, but whatever it is that you're good at, expect us to offer you a chance to show us that you're good at it.
It's probably only a useful technique at companies which are big enough and flexible enough to find a use for you, whatever your skill set happens to be. For a company that only has one job opening and needs a specific sort of person to fill it, this sort of pick-your-interview-subject approach doesn't help them.
(Trick questions are useless: they primarily serve to make the interviewer feel clever, not to demonstrate the abilities of the candidate)
Google had a list of technoliges as you described, and I was asked to rate myself on each one. I think it would have been slightly better if they explain why or how it is used.
My one experience with getting to a Google onsite is that, despite repeatedly telling various recruiters my strongest areas, they ended up interviewing me for what they wanted me to do rather than what I was good at. In fact, one of my interviewers literally had to throw out all his prearranged questions and make shit up as he went because my background and experience were so misaligned with what he had been added to my loop to ask me about.
So maybe they were right to reject someone who answered 4-6 in most categories. Either that person isn't a good fit skill-wise, or they are a good fit skill-wise, but not personality-wise.
Realistic assessment of your abilities matters as an employee.
Humility is far rarer and more valuable.
You did not have a phone screen. What you are describing is the initial contact with a recruiter, to establish what roles you're interested in, explain how the process works, and arrange dates. Had you proceeded past this, you would probably have had a phone screen next, which would be with an engineer giving you a technical problem to solve.
I also rated myself 4-6 on most categories, and was hired; that's an entirely reasonable score on the scale they give. I don't know what happened in your conversation with the recruiter, but it sounds like you made it clear you weren't interested and that was the end of it.
My not be considered a "phone screen" in the process but if it walks like a duck and quacks like a duck.. :)
Anyway, here is the cache version: http://webcache.googleusercontent.com/search?q=cache:61DmBRy...
Basically, this interview process allows you to weed out candidates who
1) disapprove open floor plan work environment, sitting next than six feet away from co-worker (yeah let's cramp twenty people together around so we can hear each other yelling at the headphone)
2) markers are either missing or used up and no one replaces them. oh you want to project your screen? good luck find a working remote or HDMI cable.