Comment by potatolicious | Hacker News Reader

Comment by potatolicious | Hacker News Reader

Parent Full thread

> "Google have [probably] rarely interviewed anyone who was not well qualified - yet around 90% of candidates fail the interview."

This statement hinges on the belief that these 90% of candidates fail because, while qualified on paper, they are not actually qualified for the job.

I suspect it isn't actually true, and that in reality the Google interview process fails very large number of actually-qualified candidates, while also admitting a large number of not-actually-qualified candidates. In short, I suspect beyond pre-screening people for paper qualifications, the actual on-sites have an effectiveness in the order of a coin toss.

But yet we take pride in failing hordes of people, because it somehow implies nice things about our surviving it. Somehow the fact that the process for hiring programmers doesn't involve any programming, and that we are hiring people to architect large systems by solving isolated CS101 algorithms questions, doesn't bother us.

It doesn't really hinge on them not being qualified for the job the way I read it. It hinges on them not being 'the most qualified' for the job. As failure of the interview to me is defined as not getting the job. And interviewing 10 people for every position leads to them choosing a most qualified out of those 10 and leaving notes for future teams looking to fill a position.

I am no way saying their practices are not discriminatory, just that I read that statement differently.

I think due to the complete lack of disclosure about why you don't get a job in the tech industry it will be interesting to see Google have to justify their decisions with the facts to a court. I personally ask after most every interview what I could do to improve, or what I did wrong. With very rare exceptions I am told they can not disclose that information for liability reasons. (Google and Amazon both definitely gave me this answer) A handful of companies have told me things like, we went with a candidate that had more experience in ________. And once a company told me that a project was cancelled so they unfortunately no longer had a position available.

> "And interviewing 10 people for every position leads to them choosing a most qualified out of those 10 and leaving notes for future teams looking to fill a position."

I'd agree with you if the above was true - but it isn't at Google, nor is it true throughout the majority of the industry.

Most companies (Google included) hire based on a fairly simple "first past the post" system. A candidate comes in, interviews, and a hire/no-hire decision is made shortly after. In cases where two candidates are interviewing simultaneously some comparison might be made between the two, but otherwise the candidates are decided upon serially.

I don't know of any companies that practice the "interview N people, choose best" methodology - if they did, then yes, I agree that "failing" 90% of people might be more reasonable.

I don't even know if that strategy is practicable - sourcing candidates generally takes a long time. Getting N people to interview within a short time window for the same position is nearly impossible for any N > 2.

The company I work for very much does that, as have others I have interviewed at and knew people working at. The company I currently work for is legally required to interview N (I don't know what N is) candidates for each position, hiring takes a long time due to clearances, and they have firm dates they need to have people started by so they can finish projects on time as they get projects. In the case they can not find enough qualified candidates where I work they relax the criteria and interview unqualified ones and hire the best they can find in that bunch, probably how I got hired with no experience. A candidate search usually starts 6 months prior to a project starting, with a decision made at least 2 months prior. While it is not the company I currently work for Salesforce.com is a company I have good reason to believe uses a system of choosing a single person out of a pool of interviewed people after a certain number have been interviewed.

Since google has the multiple rounds of interviews and they very much had notes on my first time interviewing for my second time interviewing (neither of which did I make it to the second round of in person interviews). I suspect that googles last day of in person interviews, (2nd day? 3rd day? don't know how many days of in person interviews they have but I know it is more than one) is when they perform things in the manner you are describing. I have been told all the technical portions of the interview are over by the last day of in person interviews by some people who work there, although I am taking that information on faith based off the fact they are people I trust, and do not know first hand. Also from what I could tell Google does not use a short time window for their interview process either, as each time I entered the process it took over 3 weeks to go from recruiter to the first in person round.

At growing companies like Google, they hire everyone who meets their vague "hiring bar"

The times I interviewed at google it was made clear to me that I was interviewing for a position on a specific team, and not a general hiring pool. So far as I met most of the team members, maybe you are correct though and the team size was just very flexible.

Every place I've worked, either being interviewed or doing the interviewing, has worked on a "multiple interviews and pick the best" basis.

This first past the post nonsense is crazy.

Reply on news.ycombinator.com