I'm rethinking what I did on my last hiring position. I sent a take-home test that should take about 2-3 hours to those who sent in a decent resume. It filtered out a little over half of the applicants I responded to, as in half didn't respond. It further filtered out a small segment who didn't do very well on the test. A filter is definitely needed to weed out those who are simply casting a wide net, are looking for contract work instead of a FTE and to those whose skills simply do not match what they've purported in their resume.
I really can't call everyone who has a decent resume, its way too time consuming. If we had a recruiter then maybe it would be worth it, but not personally. Maybe a take-home quiz that takes like 20-30 minutes would be a good filter instead?
>Wouldn't it be funny if the half that didn't respond were the best candidates you heard from, and you threw them away with your take home thing?
In my estimation, it's not likely that all of the ones who didn't respond were better than the ones who did, but it's likely that a sizeable portion of them were. They're not going to jump through ridiculous hoops because they don't have to.
One time an employer of mine wanted newly hired developers to sign a crazy onerous contract with unbelievably broad and vague non-compete provisions and IP provisions. I told him that the only people who would sign it were people who were desperate for employment, or people without integrity who would sign it without any intent of abiding by it. Neither of which were the types we were looking for. So he relented and we ditched it.
It's the opposite effect IMO. A good test is more likely to keep the better candidates and only reject the poor candidates.
The exercise is sent upfront. A good candidate can estimate that it's a reasonable program doable in an hour. It's up to him to decide whether he wanna continue the application or not. Either way, both parties win.
A bad candidate who can't code cannot return anything. Little risk of false positive.
I think the only important thing is to keep the exercise short enough, one or two hours top. It doesn't exclude people who have a family or other obligations.
That's almost like an exercise in economics. There's a middle ground of what the buyer (i.e. the employer) wants and what the seller (i.e. the applicant) wants. Buyer wants all sorts of qualities + a low salary + an NDA and non-compete, seller wants a high salary + a good work place + no NDA or non-compete contracts. Somewhere in the middle both buyer and seller should be satisfied, if one can live without the NDA contract then its not needed in the negotiation. If its a non-starter then the applicant can look elsewhere.
For sure, its something I've thought about too, but at the end of the day you need to do something to get from the many down to the one. Gut intuition gets used at some level and I prefer that to be way down the line after we've gathered a lot of data about a candidate. Take home tests also serve as a standardization measurement. Everyone does the same test so we can at least compare candidates to each other on that level. If they don't do the test then its pretty tough to compare them.
At some point you need to ask the applicant to do something to prove their worth and interest in the company. Take home tests are only one filter for that and I'm honestly open to others.
>I really can't call everyone who has a decent resume, its way too time consuming.
Even if you have, say, 100 resumes you consider "decent" (which to me would be an impossibly high number) then you need to be able to rank them, so you can pick, say, the top 10, and invest some followup time in those.
Totally, by the time we have a phone interview I'm pretty seriously considering that candidate and I want to get to know their personality better to see if they're a good fit, so whatever I can do to shrink that pool in a way that's fair to everyone is necessary.
I love trying to square these conversations about needing to shrink the candidate pool to something manageable with the widely claimed “shortage of engineers”.
That's easy, there is a vast chasm between people who say they can do things and people who actually can do things. This is precisely why these filters exist!
>That's easy, there is a vast chasm between people who say they can do things and people who actually can do things. This is precisely why these filters exist!
Amen. There is no shortage of applicants. But there is a shortage of competent, qualified ones. In fact, I find it frightening the degree to which blatantly incompetent people can skate by in our industry. I recently interviewed someone who rated themselves as a "professional level" Java programmer. (The highest level he could have rated himself is Expert, one above professional level.) The first technical question I asked was, "OK, please write a Hello World program in Java for me.". He totally bombed. It was painful to witness. He had no clue what he was doing, but yet, there he was, asking for a six-figure salary and deeming himself worthy of it.
Supply and demand. Whether there is an actual shortage of a good depends on price.
The extremes:
Is there a shortage of talented engineers willing to work for minimum wage? Of course!
Is there a shortage of talented engineers at a price of, say, $5M a year? I guarantee not.
So there exists a market wage somewhere in between at which supply of talented engineers must equal demand. If you think there is a shortage of real talent you might not be at market rate for that talent.
Using your example, there exists a market rate for an engineer who doesn’t know how to program Hello World. You might be surprised how much that is!
I really can't call everyone who has a decent resume, its way too time consuming. If we had a recruiter then maybe it would be worth it, but not personally. Maybe a take-home quiz that takes like 20-30 minutes would be a good filter instead?