Whats the goal of behavioral interviews in engineering ?
Do people actually give "bad" answers in behavioral interviews ? I can't imagine someone saying : "Yeah, if the project is late then it's late, what about it ?" "If I have to work on something I disagree with then I'm going to constantly bitch & moan about it or just not do anything at all, it's my way or the highway" "If I have a dispute with a colleague I generally leave on their desk telephoto pictures of their kids leaving kindergarten, that generally helps out a lot" I think there's a lot of opportunity for failing a behavioral interview just because there is no way to justify in 2-3 min how you handled making a decision in a context that took years to develop. Add on top of that communication / assumption issues on your and the interviewer's part and you have a recipe for disaster. Ex: You: "The difficulties with the project were constant redesign / shift in direction from the VP" Interviewer: "And what did you do to avoid that in the future ?" You: "There's really nothing I could do ...." Interviewer: "<Candidate shows lack of initiative, easily gives up and seems to be lacking core leadership skills>" When in fact ... there is absolutely nothing you can do realistically in the specific situation because doing anything about it would be career suicide. So, what is the goal of the behavioral interviews ? If it's "avoid hiring people you wouldn't want to work with" - I think that it's exactly those people that are the best at finding a perfect answer for a tricky behavioral question. Also, Amazon is the most behavioral-question heavy interview and from what I understand they have the biggest density of "people you don't want to work with". If you went to behavioral interview training, what red flags were you told to look for ? Have you actually heard a "bad" answer that could not have been the result of misinterpretation / assumptions / communication issues ? What answers have you failed candidates for ?
The goal is to filter out the jerks and the immature morons. Doesn’t always work but that’s I think what the goal is.
From my understanding Amazon has the biggest density of jerks - not sure about immature morons. Amazon also has the most behavioral questions (~50% for SDE2) and 60%-70%+ for senior Also from my understanding Google has the lowest density of jerks - and they have no behavioral questions for SDE2 or even Senior I think. Also Microsoft doesn't have any behavioral questions and has very few jerks from my experience, I've heard lots & lots & lots of ppl complaining about incompetent / technically bankrupt people, but very very few complain about jerks. Given the Microsoft situation - where 0% behavioral and 100% technical ends you up with a big problem in incompetent ppl and very small problem with jerks - I don't see why you would make your interview 50% focused on anti-jerk, if behavioral is actually meant to be anti jerk.
The few times I interviewed with Microsoft, a solid half of the phone screen/first round were annoying behavioral questions ("tell me about a time ____ happened") so Microsoft definitely has them.
Companies I've seen with the most behavioral - LinkedIn and Reddit (2 rounds for LinkedIn onsite, 3 for Reddit onsite) Maybe Microsoft, in the early rounds. I've heard for Amazon too.
I think behavioral interviews can be effective, but 99% of interviewers don't know how to do them properly.
I ask it a lot - just asking is of no use , interviewers need to know what to look for . What I look for is the depth of answer and variations in answers and examples - if there are few examples then its shallow
The point is to get insight as to how a candidate has behaved in common situations in the past, as that is the best indicator as to how they would behave in similar situations in the future. They have their pluses and minuses, but all in all I would much rather do a behavioral interview than play 20 questions of the sort: tell me where you see yourself in five years.
Common red flags from behavioural interviews include when asked about successes, the response includes "I did this.. I... I... I...", but when asked about challenges and failures, it's "they did this.... They... They... They..." Engineers all require the ability to be introspective, acknowledge weakness, be empathetic, have customer focus, collaborate, and require strong values. These qualities can all be teased out of behavioural interviews, it all just depends on how they're asked. You're right though... Given clinical, calculated questions, you'll received canned, calculated answers, but when the interviewer makes it seem like you're just shooting the shit, and having a great conversation, they're really doing a proper behavioural interview. The less you feel it, the better they are.
The other thing these interviews are useful for is deciding levels for candidates. How a new grad would resolve conflict should be much different than a senior staff engineer. If the expectations for a staff engineer is to have the ability to lead efforts and deal with ambiguity then the behavioral interview is one of the ways to tell if they have that experience already. If they don’t then they will likely get brought in at a lower level
Yes, people do provide answers that are mind blowingly bad. The following two are actual responses: I asked, "Why are you leaving your company?" Candidate replied, "Everyone in my group is an idiot". I said that might be true, but that's usually not the best response when asking why you leave a company. Where there other reasons I asked? . He replied, "But it's true, they are all idiots." I asked one candidate why she wanted to join Oracle. She said, "My ex lives in San Diego so I'm just trying to join any bay area company just to get away from Southern California."
So you punished them for telling the truth? If they would have lied to make general sense, would you have gotten a different candidate? Wouldn’t it still be same candidate? Unless you are saying you can see through candidate magically...
The "idiot" one is pretty bad, prob a fail, yes. If it was about the team that would be more legit since if a team is majority bad, then all good people would leave leaving it in a bad state. For a group it's more difficult - perhaps the culture in the group is very political or whatever - and everybody either adapts or falls behind. So "idiot" could mean "falling in line and refusing to fight BS". And about leaving San Diego .... I mean ... how many people switch companies because of location ... that's extremely common. Do the specific reasons for wanting to switch locations matter ? Would "wanna get out of NYC because high cost of living / transportation options" have made it any better ?