[yamazaki]

That's why I think take-home exercises are a much better way to evaluate candidates. Give them the exercise and enough time and you will figure out:

1) How good they are 2) How much they actually care about the job offer -- the more detail and the more passion they put in the exercise, the more into the job they are.

[tpxl]

Give 20 candidates take home exercises, they're all passionate and put a lot of work into it, then you'll reject 19 of them.

The input required for this type of interview is heavily skewed in favor of the company. No thanks.

[mirekrusin]

Why not drastically different approach - ie. instead of active technical interview process on company's side, just put up few dozens of open source projects/libraries that are being used inside company and leave it as technical interview sandbox.

Keep brownie points on pull requests to unlock "you gained <<passed technical interview>> badge" a'la stack overflow or do one off evaluation based on contribution when candidate applies. Stack overflow contributions can be equally taken under account to gain points and arguably they have badge system that does the job for you already.

Github could introduce something similar and wipe out "anxiety interview" completely, if they wanted to.

There is really no reason to do technical interview when you can inspect open source work. For people who don't have contributions yet, open source libraries/projects by the company should be made available.

Seems like win-win for everybody.

[eNTi]

Unless you are looking for a "senior developer with x years of experience". I have a live, hobbies and family. I don't care to spend most of my free time fixing code in some obscure library I'm not ever going to use only to stand a chance to get a different job in the future.

How in the world did current developers EVER get ANY job? Who gives developers the chance to grow these days?

[reagent_finder]

It's a "foot in" problem, I feel. Shit companies have shit interviews and give out shit workloads to MAYBE get a foot in the door.

You can follow along or be proactive, for instance start a new github account just for interviews, do code wars challenges, write a blog and just link your writeups, github and blog to applications. Along with your CV with major projects etc. of course.

Most people I know who work at big firms got calls from friends or friends of friends to interviews, so spending time socializing the local hacker circles is probably going to be worthwhile as well.

If you have life, hobbies and family, you should still be able to schedule 1-4 hours a week to job hunting, much more if you don't have a job of course.

It's not pretty or much fun, but it's certainly doable.

[reagent_finder]

Not the worst approach, and as long as it's in the spirit of cooperation might even be good.

UX people could tune or at least comment on the interface, security people could fuzz or pentest the library... however, for a lot of actual problems in actual open source libraries it requires a crapton of internalization -- again, for free. Granted, it's for open source but it could be argued it's even WORSE approach for the company, since they're getting free benefit from expertise and work.

Also, I'm a private person, I have splintered online identities, I'm not going to give my 'hobbyist' github account in a professional context, and I'm also certainly not going to make a github account just for interviews and applications.

I'm lucky enough to have connections that will probably keep me employed as long as I like, and the tales of interviews and free work that's required to put up just to get a chance seems insane.

4 hours... okay, I can sort of see that, I might set myself a limit of 1 hour and tell them this was my approach, this was how far I got. I've heard of 30-60 hour workloads given to applicants to 'top five' companies and it's just ridiculous. As are the stories of interviewers who have no technical skills themselves and are just looking for a canned response (Google, looking at you).

I was going to say that the best approach I've seen are companies that put up interesting puzzles and security challenges on their website. The technical skills required aren't _that_ high, but certainly require an amount of ingenuity, persistence and just plain being interested in tinkering that they feel suits their company profile. The applicants can solve or try to solve these challenges, do writeups and probably get interviews just based on those.

Granted, I can also understand someone looking at 50 companies all giving them 4-10 hour workloads to MAYBE get an interview might feel frustrated and overloaded. It's a classic egg-and-chicken problem. If I were to start looking for a job from scratch I'd probably solve a bunch of code wars challenges and security CTFs, make a blog and just link my writeups and blog to applications.

[legopiece]

Totally agreed. I've been in the position to evaluate several take-home exercises during the years and they always produced good results

[eNTi]

Doesn't that tell you that the method of hiring is flawed in the first place?

If you care to weed out bad candiates, what help are 100% perfectly fine take-home exercises? The only thing that tells you is, that all of your candidates are viable in the first place. Just go into an interview with THAT assumption and ask the right questions.

Why do so many companies assume that people who try to get hired for a certain job don't meet the basic requirements? That gotta be the rule not the exception? That's just a bad faith assumption.

Just make sure your hire fits into your culture and give them sufficient probation time to grow into your codebase.

[cutemonster]

> > produced good results

> what help are 100% perfectly fine take-home exercises? The only thing that tells you is, that all of your candidates are viable in the first place.

I think you might have interpreted that, in a different way than what was meant.

[reagent_finder]

Could you provide a bit more info? How long were the exercises? Were the applicants evaluated more based on their results or possible writeups? How much did their CV weigh compared to the take-home exercises?

[bananaface]

Aren't you worried about filtering out good candidates who have enough options that they won't spend time on your test?