4 ms·

Feels like we constantly talk about how interviews suck, but we kind of all know why. 1. It's tough to fire people (in that, if you ask someone to move and it

by diob 6y ago

Feels like we constantly talk about how interviews suck, but we kind of all know why. 1. It's tough to fire people (in that, if you ask someone to move and it turns out they aren't a good fit, it's kind of a jerk move to then drop them like a hot potato). 2. Performance indicators are tough. How do you know someone is doing bad vs good? 3. Keeping on a bad hire (especially one you can't identify), is damaging as heck. Best case scenario they don't get much work done, worst case scenario they suck away the time of all your good hires through bad decisions. So while I agree that interviewing sucks, I've come to terms with why it is so bad.

ghaff 6y ago
Those aren't directly reasons why interviews suck. IMO it's more that interviews, no matter how done/how well done, are a very imperfect tool. So we try to make up for quality with quantity--for the reasons you say. What you probably really want is internships and referrals from employees who have actually worked with the person, i.e. they're not just a college buddy. But you can't come close to filling every position that way.
- deckard1 6y ago
  Steve Yegge has a thing he calls the "Interview Anti-Loop" or "anti-panel". Basically, for every person that works at some company (say, Google), there is a least one set of employees (the anti-panel) that would reject that person in an interview. http://steve-yegge.blogspot.com/2008/03/get-that-job-at-google.html http://steve-yegge.blogspot.com/2008/03/get-that-job-at-goog... Filtering out obviously bad employees is really not an issue. It's the fact that you're asking a committee to do exactly what a committee does worst at: making a decision. So every interview panel collectively decides that it's best to not make the decision (i.e. reject the candidate). It's the path of least consequence for the members of the panel. If it turns out the hire was bad, they don't look like a bunch of fools. > it's highly likely that someone on the loop will be unimpressed with you, even if you are Alan Turing. I fully believe this. The only reason you will ever get hired is because the company is desperate to fill a seat at that particular moment in time. It has zero to do with technical skills, social skills, or anything. It's 100% pure luck. Almost every tech company is "hiring." These companies don't stop hiring. Hiring being the advertising of positions and the interviewing of candidates. But that does not mean those companies are hiring right now. Or even have positions available for those they are advertising. It's often a big bureaucratic ship that doesn't have the capability of even knowing what their needs are at any given time.
  - ghaff 6y ago
    Maybe. I'm not involved in a lot of hiring but my general experience is that the interview panel usually ends up at one of two positions. Everyone is sort of meh (and maybe one or two people are actively negative). Or everyone is hire this person now--maybe with some minor caveats here and there. I'm not sure I can remember a situation where there was a strong split in opinion.
- diob 6y ago
  You're kind of describing the situation we're heading towards already, as I see contract to hire as essentially an internship. And maybe it's just where I'm living, but almost everything is trending towards that. It's kind of wild. We are now putting people through an insane interview, and then also telling them we are going put them through probation. Choose one, at least that's what I would say. The referrals are a mixed bag, and I think would create a bit of a revolving door where it's hard to enter the industry unless you have connections.
keithnz 6y ago
But is the "process" making it better at finding good hires? I know for myself I did big interviews at one stage, and realized most of it was redundant, I wasn't really learning more that would change my decision, I quickly learnt most of what I needed with a conversation and writing a small amount of code ( not tricky code either ). Students are a bit harder, but we participate in a summer intern program, and give most anyone a chance . We have taken on a few people from that, and that's worked well.
jennyyang 6y ago
If these ridiculous interviews don't produce top notch performance reviews after a year on the job, then the interviews are worthless. If I were to ever try a startup (I won't), my philosophy would be to hire easily and fire easily. Do my level best to interview fairly, and give people a chance. But if they don't work out, fire them quickly and give them a 2-3 month severance bonus. Then I would remove most titles and pay in the top tier. I think at the beginning of hiring, it would probably have a high turnover, but as the company matures, if it survives it will be filled with a lot of happy engineers that won't want to leave.
- diob 6y ago
  Let me know if you do that thing you said you won't, I'd be interested :)

by diob 6y ago

ghaff 6y ago

Those aren't directly reasons why interviews suck. IMO it's more that interviews, no matter how done/how well done, are a very imperfect tool. So we try to make up for quality with quantity--for the reasons you say. What you probably really want is internships and referrals from employees who have actually worked with the person, i.e. they're not just a college buddy. But you can't come close to filling every position that way.

deckard1 6y ago
Steve Yegge has a thing he calls the "Interview Anti-Loop" or "anti-panel". Basically, for every person that works at some company (say, Google), there is a least one set of employees (the anti-panel) that would reject that person in an interview. http://steve-yegge.blogspot.com/2008/03/get-that-job-at-google.html http://steve-yegge.blogspot.com/2008/03/get-that-job-at-goog... Filtering out obviously bad employees is really not an issue. It's the fact that you're asking a committee to do exactly what a committee does worst at: making a decision. So every interview panel collectively decides that it's best to not make the decision (i.e. reject the candidate). It's the path of least consequence for the members of the panel. If it turns out the hire was bad, they don't look like a bunch of fools. > it's highly likely that someone on the loop will be unimpressed with you, even if you are Alan Turing. I fully believe this. The only reason you will ever get hired is because the company is desperate to fill a seat at that particular moment in time. It has zero to do with technical skills, social skills, or anything. It's 100% pure luck. Almost every tech company is "hiring." These companies don't stop hiring. Hiring being the advertising of positions and the interviewing of candidates. But that does not mean those companies are hiring right now. Or even have positions available for those they are advertising. It's often a big bureaucratic ship that doesn't have the capability of even knowing what their needs are at any given time.
- ghaff 6y ago
  Maybe. I'm not involved in a lot of hiring but my general experience is that the interview panel usually ends up at one of two positions. Everyone is sort of meh (and maybe one or two people are actively negative). Or everyone is hire this person now--maybe with some minor caveats here and there. I'm not sure I can remember a situation where there was a strong split in opinion.
diob 6y ago
You're kind of describing the situation we're heading towards already, as I see contract to hire as essentially an internship. And maybe it's just where I'm living, but almost everything is trending towards that. It's kind of wild. We are now putting people through an insane interview, and then also telling them we are going put them through probation. Choose one, at least that's what I would say. The referrals are a mixed bag, and I think would create a bit of a revolving door where it's hard to enter the industry unless you have connections.

keithnz 6y ago

But is the "process" making it better at finding good hires? I know for myself I did big interviews at one stage, and realized most of it was redundant, I wasn't really learning more that would change my decision, I quickly learnt most of what I needed with a conversation and writing a small amount of code ( not tricky code either ). Students are a bit harder, but we participate in a summer intern program, and give most anyone a chance . We have taken on a few people from that, and that's worked well.

jennyyang 6y ago

If these ridiculous interviews don't produce top notch performance reviews after a year on the job, then the interviews are worthless. If I were to ever try a startup (I won't), my philosophy would be to hire easily and fire easily. Do my level best to interview fairly, and give people a chance. But if they don't work out, fire them quickly and give them a 2-3 month severance bonus. Then I would remove most titles and pay in the top tier. I think at the beginning of hiring, it would probably have a high turnover, but as the company matures, if it survives it will be filled with a lot of happy engineers that won't want to leave.

diob 6y ago
Let me know if you do that thing you said you won't, I'd be interested :)