Hacker News Reader

Comment by sho | Hacker News Reader

> One year, I hired 50 engineers. I couldn’t have done that in a high quality way without a system that let me interview about 500 candidates

How is this even physically possible? I don't mind interviewing usually but 2 or more a day, every single day, sounds like pure hell. Not to mention that after not touching any code for a whole year I'd be questioning your ability to be doing technical hires in the first place.

I'm very skeptical also that this ultra-standardised approach is getting anything like what I would call "top 10%" programmers. Top 10% of people who happen to do well in this particular interview setup, morelike. Teams I like to work for have a healthy mix of abilities, passions and experience. This sounds like a recipe for a room full of dull clones, none of whom have any knowledge outside a strictly specified little box. What next, psychometric screening?

I do agree with one point though - 3 people in the room is probably the right number.

View on HN

> What next, psychometric screening?

I did one about 10 years ago, and let me share a funny story. The head of HR was sitting next to me and right after the psychometric test she gave me a blank sheet of paper and told me to draw a human being. I promptly asked (without even considering my question, I was ready to ask the person's gender right next, looking for requirements): with clothes on or naked? She laughed so hard that I thought I had failed and laughed at it too. I got the job, though.

That would have been an absolute perfect opportunity to to draw a middle finger and walk out. That sounds like an absolutely idotic application process.

I might have even laughed at the HR person and asked the question "are you fucking serious?"

Agreed. That's beyond the pale and simply unprofessional.

Sounds like you've got a fair bit of experience limiting your own career.

Not really, a company that's employing those tactics is probably one not worth working for. An interview is a two way street, they are interviewing me and I am interviewing them. This would be a huge red flag.

The best companies I've worked for have been the ones who put a lot of effort into making sure they understand who they're hiring. That usually means a lot of weird questions about topics that aren't related to the technical aspects of the job.

My last round of interviews I got the tech side without any issues, but I was subjected to three rounds of the soft skills side because they were concerned my quiet affect might not fit into the team. I did what I was asked and was rewarded with working on a fantastic team that someone had put significant effort into ensuring the personalities meshed well.

Burning a bridge because you thought someone cared too much about the wrong thing is a great way to never get another callback from everyone involved in the process. Recruiters and HR people move around a lot, and have long memories.

I'd rather someone cared too much, than not enough.

I'd happily burn a bridge with someone in HR that expected me to jump through hoops and draw pictures. Those people need to hear the truth, they are bad at their jobs.

> How is this even physically possible?

He is Director Of Engineering. A large part of his job is supposed to be hiring if they are growing.

> This sounds like a recipe for a room full of dull clones

Can you explain why? I've found that a standardized interview with asking the same questions to each candidate helps me better see the differences between candidates.

Over the years I've tried on a lot of different interviewing processes, and what the article described is pretty close to what I've ended up with as producing the best results.

> I've found that a standardized interview with asking the same questions to each candidate helps me better see the differences between candidates

Oh, I start with a standard list of topics, and of course technical questions need to be consistent across interviews to be any use. But beyond that I like to let the interview flow naturally - if the candidate seems particular passionate about one thing or another, follow up on that, and see where it goes. It's a conversation, not some form to be filled out. If you're going to do that, may as well just email them the actual form.

Maybe the author isn't as strict as I understood but I really do believe that it's in that conversation that you get to know someone's personality, what they're passionate about, how they might fit culturally and how their experience and talents might complement the team. You don't get that from a survey.

I agree with you about letting the interview flow naturally. That's how the last half of my interviews go. Often times looking at and discussing some of the candidate's code, depending on what the role is.

I don't always get it perfectly right, but when I do my interviews feel good to both sides. It just feels like we spent an hour discussing a topic we both like and we both learned something new.

> Maybe the author isn't as strict as I understood

I didn't get the idea that he is when I read it. What I did get is that he wants lots of data points confirming the same thing. A list of standardized questions or topics is just one of those data points, right?

> Not to mention that after not touching any code for a whole year I'd be questioning your ability to be doing technical hires in the first place.

I you read code from 20 years ago, you will be surprised that there is good and bad code. Most of the code techniques were developed decades ago.

If you stop coding you are going to be more slow when you have to start to do it again. But you don't forget the principles.

You will find managers that user to be developers that are bad coders. That does not mean that they have forgotten how to code, in my experience they were not good to start with and that is why they changed profession.

> Not to mention that after not touching any code for a whole year I'd be questioning your ability to be doing technical hires in the first place.

Let's not get carried away. Programming isn't easy, but are you really saying that after 1 year not coding, you can't do technical work? 1 year is maternity/paternity leave in some countries, do you really think that people coming back from maternity/paternity leave can't code any more?

Psychometric screening is already used in some places actually.

Yeah, I've heard that. I'd simply leave if any company asked me to take one of them. You couldn't find a bigger red flag in China.

Amazon does it

Although I wouldn’t work for a company which blindly made hiring decisions based on a profile, the approach in this blog post would make the company more attractive: https://www.xaprb.com/blog/personality-assessments-hiring/

Excerpts:

> The assessments are never really correct, so a vital step is reviewing the report with the candidate

…

> How the candidate reacts to the report; which things they think are true or false and why; which things they engage with and which they skip over; those are all revealing.

I cofounded a company last year, and our focus is on facilitating 360 reviews / annual reviews for companies. The format of our assessments are fairly “psychometric”, but not because we use it to profile someone, but because it helps facilitate written and verbal feedback from your peers and manager. The excerpts you highlighted match our findings, and is a big reason why we run workshops as a part of our process (so we can explain how the information is used).

This is the same way I approached technical interviews when I was doing a lot of hiring. I didn’t usually care a lot about how someone answered a question, but rather how they reasoned about that answer.

Although not in the same "domain", I have had good results using viainstitute.org's character strengths tests to better understand my professional and academic clients before creating/editing their resumes, linkedin profiles and interview coaching. Its been around awhile and focuses on what people should do more of. YMMV.

Thanks for the link. That does look like a much more thoughtful use of the technique than I had anticipated - talking about the result afterwards is a good idea. I had been thinking their more common use was indeed to blindly discard candidates based on some nonsense quiz, totally ignoring the creativity and passion the candidate could have otherwise brought to the team.

I've actually done a couple of these tests, although for the life of me I can't remember in what context, and thought them ridiculously simplistic and obviously gameable. Now I'm curious to try some of the more subtle ones the author alludes to.

I don't understand. Is the term psychometric screening here being used to refer to personality / aptitude tests? I think I'm confused because this is not how the term is used in psychology, but maybe businesses use it this way?

Hard to argue with the results. He ran engineering at Clover Health from the very beginning and it’s now a highly successful company.

I think "standardising the process" doesn't mean "multiple choice test". On the contrary i had the impression the author is putting much care in getting everyone to know the human qualities of a candidate.

But you have to define process if you want to tie in your team and consider many candidates.

> What next, psychometric screening?

I actually did one during the interview process for a Toronto-based startup.

Reply on news.ycombinator.com