11–20 of 21 posts
It's similar to making a decision to fire someone: if you have a suspicion that someone won't work out, probably best to just immediately cut them. Don't think I've ever seen anyone in my entire career come back from poor performance unless it's due to unique / temporary life circumstances.
When candidates are clearly unable to finish the interview problems even with hints and help, I conclude the technical interview early and transition to an informal discussion of their career, progress, and goals. I don’t think it’s worth compromising your technical interview standards just because you like a candidate. You have those standards for a reason. However, having a friendly conversation with the candidate…
That said, I've considered doing what you've suggested, but I worry that comes off too obviously as having given up. I also can't always speak to the expectations of the particular role because it's often not on my team (or department in some cases). There's been times where I have cut the interview short and done something similar, but I'm hesitant to try that on a particularly bad candidate, because I worry I wouldn't be able to stay away from asking how much of their resume is a lie.
Just end the interview early: why waste their time and yours? It's similar to making a decision to fire someone: if you have a suspicion that someone won't work out, probably best to just immediately cut them. Don't think I've ever seen anyone in my entire career come back from poor performance unless it's due to unique / temporary life circumstances.
It has happened to me. Sometimes you are tired at times the question/solution doesn't click immediately. At those times I wish the interviewer would leave for a few minutes so I could focus completely and grasp the problem and get started. The observer effect of watching and trying to hold a conversation breaks focus.
The goal of an interview is simply to identify the skills / personality of applicant. I have a role in mind when I am interviewing someone, but if their skills do not support this but it feels like there may be a good match at another team (at the right price) then I start feeling them out to see if that could be a match for them. If their price is too high for their skills and/or I do not have a match, I politely bu…
the goal of an interview, apparently, is to see to what extent the applicant was willing to repeatedly grind the same leetcode problems over and over again and memorize their solutions
If it's important to you to find people who can write code, why are you even interviewing people who "explicitly say they don't know how to write code"? And if you interview them, why are you interviewing them in the same way as people who can write code? And if you interview them, why are you keeping them for 45 minutes on the struggle bus instead of interviewing in a manner they might actually succeed in?
Earlier quoted context omitted.
the goal of an interview, apparently, is to see to what extent the applicant was willing to repeatedly grind the same leetcode problems over and over again and memorize their solutions
I would never ask leetcode questions that have some trick to them. These are ones we've come up with internally and are incredibly straightforward.
I think that's a sign right there that you need a phone screening stage before you bring a candidate in and tie up a whole day of your team's time.
It can be an informal 15-30 minute conversation but it should be someone technical performing the screen, not an HR type. Take 5 minutes to explain the role. Then confirm the matching capabilities on their resume and previous jobs and begin to ask some precise questions about how they used them. You can even ask some simple questions. What's a join? What does vacuum do?
You're not asking brain teasers but you're putting a very heavy bullshit filter on their responses. It shouldn't be hard to figure out who is really capable and those that have never written a lick of code. You'll also catch people that can't communicate technical ideas to a stranger, which is typically a red flag for me as well.