When I’m given a resume with GitHub links, I jump at the opportunity to check them out. (It almost never happens.)
You don’t?
You don’t?
Anyway, there is no reason to spend the time looking at what people have done on GitHub. Once you make the final five or ten candidates, we could honestly use sheer chance to decide and end up just fine. The only reason we don’t do that, is because it would be disrespectful to the applicants, but the interviews are as much an opportunity for them to figure out if they want to work with us once they’ve met us, as it is for us to do the same.
Hiring people once you move past a certain skill level is much more about finding a common fit, so that you can continue having the good culture that you have already established.
It’s not that I think you are wrong to look at github though. It’s just that I don’t feel it’s really ever worth the resources you need to allocate to get a proper picture. I feel the same way about the coding interview, it’s just a waste or everyone’s time. I mean, how can you tell if their GitHub profiles aren’t doctored? If it was too clean or too good, I’d personally be suspicious why they’d work for us.
It's just a little attitude thing that makes so much difference. The way salaries work, they probably don't get 5x salary so they are extremely valuable for the company and to find in general.
If you just look at general performance, resumes and things like that you might just get "rest and vest" types.
Really? Because that’s not what our data shows on it. It shows specifically that regardless of what we submit applicants to won’t matter. Maybe we’re just bad at hiring, but with 25 years of data from a very broad range of job types (public sector) we can at least be comfortable with the knowledge that everyone else sucks as much as us.
I get your point though. Some engineers are more valuable than others, but the thing is, that it’s not a constant and there are no real way to make sure you both attract and keep them.
We employ an engineer who just might be the countries leading techie on ADFS and how it plays into our national and EU vases strategies for NSIS certifiably authentication. With a background as a bouncer, I’m not sure how many places would have guess that when he was first hired.
I’m another such story though in a different manner. I used to be real rockstar developer, and a real workhorse. This was way before I got into management and long before I had children. Because when I did have children it turned out that may undiagnosed ADHD could no longer fit into the responsibilities of adult life and I worked myself into first a nice range of stress, then anxiety and then a depression. Now I never work more than 37 hours a week unless there is a big event, like elections, and when there is, I make sure to not-work the extra hours after wards. If you had hired me the year before my first daughter was born, you would have gotten the workhorse with a very high degree of both creativity and getting things to actually do something useful for the business end. A year later you would have been sitting with a depression stricken employee who was on partial sick-leave for 9 months.
I understand the dream of course, but unless you can show me some data on someone who figured out how to actually purposefully hire the dream employee, I’m personally going to consider it a dream. An unhealthy one even, because almost nobody is ever really irreplaceable in an enterprise organisation. Sure the loss of some employees are felt harder than others, but the truth is that IBM could stop selling consumer PCs and still trundle on. So in my opinion it’s much better to create a team of people who work well together and who have a good culture, because that means it’s also easier when someone moves on to other things. It also means that you’re not as effected by the life changes of your employees.
Ask yourself this when hiring: you genuinely have a series of actually-actionable tasks ready to go for the new hire? Do you have the right resources on hand who understand that a part of their job is to answer questions about those projects from the new hire (and it will probably be a whole lot of stuff)?
The answer is going to be that you probably don't, and so whoever you hire, no matter how good they are, your company will just never know.
In my experience, it's pretty simple. Pay them lots of $$$. My circle of friends is in their 40s now, and the ones we all generally consider really good make boatloads. It's hard for them to switch jobs because most places aren't willing to pay a "regular developer" nearly as much.
I never send my GitHub links either. If I am coding a side project and I think it has business potential I am unlikely to start with open source and if it doesn't have business potential, I wouldn't bother. Also if I am building something on the side, code is likely to be rushed and seemingly horrible.
I have no idea what in GitHub was written by the owner itself, what was copied, how long it took them etc.