For what it's worth, one of the things I did with my interview process was to make all the coding involve an actual computer. I gave the interviewee their choice of editor, and was happy to let them use their own computer.

My theory was that in an interview I wanted to see people at their best. I wasn't hiring them to code on whiteboards or learn how to use new text editors; I was hiring them to spend a lot of time coding in an environment they liked.

More details here: http://williampietri.com/writing/2015/slightly-less-awful-hi...

On the flip side, as an interviewer I give all candidates the opportunity to use a text editor, and less than 15% of candidates take it. The rest opt for whiteboard.

Lucky you, I was once asked to write code on a napkin and on another occasion to say lines of code over the phone :)

Well, all it really shows is that the "market" in this instance is incredibly inefficient. People need jobs, and if the vast majority of companies have a poor interview process, and 1% have a good one, assuming that this is the only thing that people care about (it's not), all that's going to happen is the 1% get the employees they want, and the other 99% still get some employees.

The market is likely inefficient for the same reason that the advertising industry is inefficient - difficulty to show than any given change independently improves outcomes.

For some jobs, sure. But what I really think it suggests is a power asymmetry. A company can do without 1 worker much more easily than a worker can do without 1 job. Plus, current processes let hiring managers feel powerful. Given the high percentage of hiring managers who are social primates, I expect that's a big factor.

In tech, though, you hear endless complaints about how hard it is to hire, so I didn't hesitate to optimize the process for interviewee happiness. I expect it would make people more likely to apply, more likely to stay through the process, more likely to accept an offer, and more likely to refer friends.

Regardless, there's enormous waste to be removed from the process for anybody who's willing to put a little thought and work in.

Google, Facebook, Amazon are all a different category than most other companies that are hiring software developers. They care a lot less about false negatives because their hiring pool is so large. They are prestigious places that a lot of people really want to work at. As a result they have more candidates than they can reasonably get through without some massive filtering.

Some of that filtering, whether intentional or not, is how rigorous the interview process is. They can afford to be that rigorous though because the fact is that even if they miss out on one really good candidate there are 10 or more replacements in the pipeline.

The same does not apply for smaller companies with less engineering centric workplaces. If they try to do it Google's way they'll be shooting themselves in the foot.

> "still have more applicants than jobs"

This is true for a lot of companies - but there's only a surplus of labor if the applicants are qualified.

For a company to be able to be frivolous with their process, they'd need to have more qualified applicants than jobs, and I think even for many name-brand BigCorps, this isn't the case.

This has been a deepening mystery throughout my career. I see most companies are utterly uninterested in improving their recruitment processes, and are at the same time frustrated that most projects fail or don't quite deliver. Meanwhile, the best engineers I know have to look for a really long time to find suitable jobs.

There are many different constraints here. Many companies that hire software developers aren't necessarily in the "business" of software which colors their approach to hiring. They aren't experts in hiring developers so they tend to outsource it. As someone who is part of hiring decisions at just such a company I regularly see things like:

1. Pushback on work sample tests from the recruiting companies.

2. Upper management wondering why it's so hard to find candidates that can pass our work sample tests.

3. Suggestions that maybe we should be using an online coding test. (multiple choice questions of poor quality)

It can be really difficult for companies that don't have a lot of in-house experience here to figure out how to hire good developers. Much of what is known in Silicon Valley style companies isn't really known by HR in a smaller company in the midwest. They don't typically read HN or follow the tech blogs. And unless they luck into hiring a vocal Senior Developer like me to specifically guide their recruiting and interviewing process they won't even realize how badly their hiring process is.

But then, apparently, there are virtually no good companies to work for. So most people shouldn't bother worrying about it.

At the company I work for, most of the employers were hired based on the strength of referrals from existing employees. The very first employees were people that the founders knew personally and thought highly of. I never had an interview because several of the employees had worked with me before and knew me. My first contact with the company was when they told me they wanted to hire me and extended an offer. I was very surprised by that because they hadn't even seen my resume or anything. That method of hiring seems to have worked out pretty well so far for employee hiring and retention. The employees we've been getting seem to be quite competent and stick around.

At the time I was interviewing with another company who seemed interested in me, but expressed a deep suspicion of new hires after having been burned in the past, and told me outright that they were skeptical of me, and that I could very well turn out to be lazy and unproductive. After three interviews in three weeks, they still hadn't made any offer. So when another company comes along unsolicited and asks me to join them because they know I'm good at what I do, I took their offer. When a company demonstrates that they value you, you're going want to work for them instead of the one that doesn't.

There are a few people at my current company that were hired with an interview based on the company needing skills that weren't available from referrals from employees, but that's uncommon. I don't even know if our company has any sort of standard interview process.

Yes and no. I have a large contingent of coworkers from another local firm that is shedding workers. As these are usually referrals, they either get a softball process or a rubber stamp.

>Or perhaps by actually reading resumes instead of keyword-filtering them

This one is a two-way street. I launched a resume writing side-business a couple years ago (resumeraiders.com) because I saw how poor many developer resumes were and how they were getting ripped off by resume writers who had no experience in hiring or in tech. Most were just failed writers.

If you expect a recruiter to read a poorly-written 5 page resume when the yes/no interview decision could have easily been made based on a summary and a few decent bullet points, that's not on the recruiter. If you think the 5 page resume is some kind of anomaly, ask the recruiters at your company how common they are.

> "Throwing darts at resumes"

That reminds me of another technique - that is you select roughly half of the resumes from the stack and bin it. Nobody wants to work with unlucky people ;)