Well, I know Googlers pretty well. Even internally, they say their hiring process really sucks. Not one person told me this, 80% of the people I talked to told me the same thing. I personally wont blame the interviewers or the recruiters, they are doing their job. I blame their messed up Hiring Committee. I don't understand why the hiring committee can't collaborate with the interviewers and hiring manager to hire a person. Everything is done through writing, and the hiring committee does everything quietly on their own. I even wonder, is the hiring committee a computer?
That's done to eliminate bias. Remember how a couple days ago, there was a story on HN about how interviewing really sucks, because ultimately, what determines whether you get the job or not whether you happen to have a personal rapport with your interviewer? People end up caring more about how you present yourself socially than what you know.
The hiring committee is Google's answer to that. The idea is to completely divorce the hiring decision from people who have personally known the candidate. Instead, there's a very broad written information channel between interviewers, references, recruiters, and the hiring committee. But it remains written, so that all your unconscious biases about people stay out of the decision, and at least in theory, it's all based on hard data.
But this an overly-academic approach to the problem. Your interviewers will still write in whatever they write, as colored by their perceptions and rapport with you. If they write that you had a bad interview, then that's going to poison the HC. There's no way around this.
The HC appears to serve only as a layer of noise/a random selection filter to determine who is actually offered a job.
It forces you to back up your opinions with data, though.
As long as the decision is made by humans, you will never get a completely unbiased result, because humans have biases. It's the same in every field - people pretend science is objective because it relies on data, but if you read Kuhn, you'll see that a lot of science is personality cults and subjective opinions and schools of thought. People pretend Google's search algorithms are objective because they rely strictly on numbers and data, but they're written by humans, and humans choose which data is important.
But that doesn't make the data useless. The act of being forced to support your opinions with data makes you dig much deeper into them, and surfaces relevant information that'd otherwise be ignored immediately because it doesn't fit your preconceptions.
Think about science vs. polemics. If you're actually doing original research in a scientific field, you'll realize that there's a lot of uncertainty hidden behind "the scientific consensus", and a lot of other ways of interpreting that same data. But that doesn't mean that the scientific consensus is wrong. It may be wrong, but it's probably less wrong than whatever vitriol Ann Coulter or Glenn Beck or Michael Moore is spouting at the moment.
That's done to eliminate bias... But it remains written, so that all your unconscious biases about people stay out of the decision, and at least in theory, it's all based on hard data.
Impossible. The interviewers, just like us, are human. They come with a bunch of built-in biases that will show through in any contact they have with the hiring committee, whether it's verbal or written.
I just don't see your "theory" holding up in reality.
And I don't see this as a bad thing. As someone making a hiring decision, I want the emotion. I want to personally talk to the interviewer and get full impression of what they thought of the candidate. You're not going to get that through an email. Technical competence is only one part of the equation. If one of this person's potential teammates doesn't feel good about the prospect of working with the candidate, I want to know that.
Of course, Google has this ridiculous idea that they can hire people and only later decide what they'll be working on and which team they'll be working in. So you end up hiring people who don't even meet their team until after they start. I just can't imagine working for a company like that -- from the perspective of the existing employee, even.
Did you read my response to gergles? No, you'll never get a perfectly unbiased process. But having to justify your opinions with written data forces you to delve a little deeper than you otherwise would, and expose facts that might qualify or disqualify a candidate, not just impressions. It's still wrong, but it's less wrong.
Culture fit is one of the dimensions that candidates are evaluated on. But again, feedback needs to have supporting data. If you didn't like a candidate, you have to say "Didn't have a firm handshake" vs. "Appeared combative and arrogant when faced with a problem he couldn't solve; said 'Google's interview process is a bunch of technical trivia bullshit' before stomping out of the room", so that the hiring committee can judge whether this is just a nit specific to the interviewer or an actual problem that will impact the candidate's prospective teammates.
The reason why new hires don't know their teams is because of Google's corporate secrecy policies: most of the time, the projects they're working on haven't launched, and therefore can't be talked about outside the company. Even then, there are exceptions. When I was hired, I was told which department I'd be working in before I signed my offer letter, who my manager and his manager were, and was given the option of working elsewhere in the company if I really disliked that. My teammate actually negotiated the specific project he'd be working on, and they created the project for him. You need to have negotiating leverage to do this, though: they not only have to want you enough for you to get hired, they have to want you more than the vast majority of other hires, enough that they're willing to bend the rules a bit.
The recruiting/interview process at Google has a lot of problems, but I don't think that either the interviewer/HC split or the whole-company teams-later approach are part of them. Both of these were created to solve specific problems that have crept into the hiring practices of other big companies, and by and large, do a better job at it than the alternatives.
What about the bias that comes from having worked with someone for a few years, and knowing without any doubt that they are awesomely gifted, loyal to a fault, and willing to dive into the breach for the common cause? You cannot figure that out with a sterile interviewing process.
A company does not have to guard against 'bias' other than racial, religious and that sort of thing. 'Bias' towards (or against) people you know is called knowledge.
I had an awesome hiring experience. Extremely responsive recruiter (I understand that's not always the case given the number of temps). Took me two weeks from on site to getting an offer.
Everyone I referred so far had a similar experience.
Same thing happened to me. I contribute to many projects, I even spent my time for some Google projects. A lot of time actually. I had an offer with another great company, and they told me to let them extend it for another week. I did the interview process at Google, and did two phone interviews and one onsite. I thought I did pretty well, but I was waiting for their reply back for a week, and no reply. Then I pinged them, they apologized and said they are waiting for references. I pinged them again for another week, no reply yet. Then I accepted the other job because Google's Hiring process sucks so much, that it tells you right there that they don't care about their employees. I spent time out of my own work life to do interviews, and they can't bother to email us in the recommended timeframe. They said within 4 days, but 14 days ... Working for a company like that would seem disastrous.
There's a consistency in these stories from different people over the years that tells me this is one of the normal processes at Google. It may not be intentional, but it is known and there is no sincere desire to fix the process. That that is so tells me the process is about asserting power and place, and not about looking for talent.
Well, why would they? Even today, they still have a reputation as the dream place to work (even tho' with what, 20,000 engineers now, they are more like HP or IBM than they are a startup). They have zero need to change until their deluge of candidates starts to dry up.
The original article makes a good case about why they need to change. They have had some serious misses lately, and some of that may be attributed to a broken hiring process that emphasizes coding ability above all other considerations.
