Resources on how expensive hiring programmers is
June 12, 2015 11:01 AM Subscribe
Just want some quick stats, such as how many hours it takes to find a good hire. Google isn't helping
I want some quick stats about how difficult it is to hire programmers. How many hours does it take to hire each engineer? How many false positive hires are there? How often are offers made and rejected?
Did some Google sleuthing and couldn't find anything on the subject. Any ideas about good places to look.
I want some quick stats about how difficult it is to hire programmers. How many hours does it take to hire each engineer? How many false positive hires are there? How often are offers made and rejected?
Did some Google sleuthing and couldn't find anything on the subject. Any ideas about good places to look.
This is like asking "how many hours does it take to organize a party"...it depends on the specifics. That said, recruiters routinely charge one month's salary for candidates who are hired, which suggests the overall cost of hiring a programmer is typically more than 1 month of effort. Joel on Software has advice for hiring programmers.
posted by sninctown at 11:14 AM on June 12, 2015 [1 favorite]
posted by sninctown at 11:14 AM on June 12, 2015 [1 favorite]
This is going to vary enormously based on where you are trying to hire the programmers, and what you are hiring them to do. It will also vary enormously based on your hiring process.
posted by aubilenon at 11:16 AM on June 12, 2015 [1 favorite]
posted by aubilenon at 11:16 AM on June 12, 2015 [1 favorite]
Response by poster: EDIT: To be more clear, I'm looking to make an infographic, so I'm fine with aggregating. The end result would be something that looks like this: http://press.indeed.com/wp-content/uploads/2015/01/Time-to-fill-jobs-in-the-US.pdf
posted by earlsofsandwich at 11:46 AM on June 12, 2015
posted by earlsofsandwich at 11:46 AM on June 12, 2015
In Silicon Valley, a standard recruiting fee is 20-25% of the first year of salary. Plus the staff costs spent to court and interview - total recruiting spend for a new hire can range from $20k-$50k.
posted by amaire at 1:28 PM on June 12, 2015
posted by amaire at 1:28 PM on June 12, 2015
Recruiting firms fee is around a months salary so looking at a salary chart for different areas may allow you to derive very rough estimates. But as saeculorum suggests it's very non-published contractual values.
posted by sammyo at 3:26 PM on June 12, 2015
posted by sammyo at 3:26 PM on June 12, 2015
Best answer: The way I diagram it out to students is a pipes and filters model. You have some base pool of candidates in the world, and they go through a set of screens:
0. The universe of candidates, qualified and unqualified
1. You advertise the role, and the candidate sees it
2. Candidates choose to apply
3. Initial screening of applications for minimum requirements, eligibility, etc.
4. Phone screen
5. In person interview
6. Offer made
7. Offer accepted, candidate hired.
I don't have solid industry research numbers on this, and suspect they're highly tuned for process, firm size, and skill set required. But I can work you through our numbers on hiring student employees:
0->1. No idea, we don't track this. Count: ???
1->2. At the end of stage 2, we have our first data point. Count: 7.5 candidates apply.
2->3. Applicants that don't meet our student employee eligibility criteria are necessarily filtered out. This runs about 20 percent, because we don't hire grad students into positions designed for ugrads. Count: 6
3->4. About 50 percent of people don't pass our online screen. About 10 percent never even start! Count: 3.0
4->5. Usually we're happy to make offers to around 70 percent of in person interviews. Count: 2.10
5->6. Industry standard is you have 2 or 3 fallback candidates per offer. We run on the low side, since we usually hire in batches, and experience low rejection rates. Count: 1.05
6->7. It's very rare for people to reject our offer. Maybe 5 percent. Count: 1.00
For validation, I looked at our last round of hiring. Last month we hired 3 developers, and had a total of 23 candidates send in resumes. So for every 1 candidate hired, we had just over 7.5 apply.
But my data set is highly specialized. For example, in industry, I figure you should expect about 50 percent of offers to be rejected, because candidates should be seeking to get 2-3 offers in order to avoid lowball offers. I make this point explicit when I diagram out hiring pipelines to students to show that if they want a high paying tech job they need multiple competing offers, and thus need to apply to at least 23 job postings, probably more.
In our case, university requires us to limit how much we pay students. Those who want to make more money can just quit school and take a job elsewhere, so we compete on intangibles like very flexible schedules, resume building, and mentoring. Our student employees are often club leaders, and are often heavily recruited by startups. We're compatible with summer internships, and students often recruit coworkers to last year's internship program.
We also require proof of enrollment. As a result, I've never had a candidate fail FizzBuzz. We stopped asking as it has no distinguishing power for us. I suspect a lot of people who fail FizzBuzz have either bought their way through school or are lying about it. Maybe both.
To go from pipeline throughput to recruiting costs, you simply add up the expenses per candidate in each round. This will vary based on how specialized your process is, how much you pay employees who participate in hiring, and how cost effective your filters are. For example, our 4 person interview panels are typically half student employees. That's about 500 dollars saved per interview.
One thing I'd like to do is mine our online screening tool and see which questions effectively predict interview outcomes, so we can save ourselves a few interviews, and maybe remove ineffective ones. I do like the idea of using our screening tool as a motivated self-training session, but we do have the 10 percent dropout rate to consider.
posted by pwnguin at 7:46 PM on June 13, 2015 [1 favorite]
0. The universe of candidates, qualified and unqualified
1. You advertise the role, and the candidate sees it
2. Candidates choose to apply
3. Initial screening of applications for minimum requirements, eligibility, etc.
4. Phone screen
5. In person interview
6. Offer made
7. Offer accepted, candidate hired.
I don't have solid industry research numbers on this, and suspect they're highly tuned for process, firm size, and skill set required. But I can work you through our numbers on hiring student employees:
0->1. No idea, we don't track this. Count: ???
1->2. At the end of stage 2, we have our first data point. Count: 7.5 candidates apply.
2->3. Applicants that don't meet our student employee eligibility criteria are necessarily filtered out. This runs about 20 percent, because we don't hire grad students into positions designed for ugrads. Count: 6
3->4. About 50 percent of people don't pass our online screen. About 10 percent never even start! Count: 3.0
4->5. Usually we're happy to make offers to around 70 percent of in person interviews. Count: 2.10
5->6. Industry standard is you have 2 or 3 fallback candidates per offer. We run on the low side, since we usually hire in batches, and experience low rejection rates. Count: 1.05
6->7. It's very rare for people to reject our offer. Maybe 5 percent. Count: 1.00
For validation, I looked at our last round of hiring. Last month we hired 3 developers, and had a total of 23 candidates send in resumes. So for every 1 candidate hired, we had just over 7.5 apply.
But my data set is highly specialized. For example, in industry, I figure you should expect about 50 percent of offers to be rejected, because candidates should be seeking to get 2-3 offers in order to avoid lowball offers. I make this point explicit when I diagram out hiring pipelines to students to show that if they want a high paying tech job they need multiple competing offers, and thus need to apply to at least 23 job postings, probably more.
In our case, university requires us to limit how much we pay students. Those who want to make more money can just quit school and take a job elsewhere, so we compete on intangibles like very flexible schedules, resume building, and mentoring. Our student employees are often club leaders, and are often heavily recruited by startups. We're compatible with summer internships, and students often recruit coworkers to last year's internship program.
We also require proof of enrollment. As a result, I've never had a candidate fail FizzBuzz. We stopped asking as it has no distinguishing power for us. I suspect a lot of people who fail FizzBuzz have either bought their way through school or are lying about it. Maybe both.
To go from pipeline throughput to recruiting costs, you simply add up the expenses per candidate in each round. This will vary based on how specialized your process is, how much you pay employees who participate in hiring, and how cost effective your filters are. For example, our 4 person interview panels are typically half student employees. That's about 500 dollars saved per interview.
One thing I'd like to do is mine our online screening tool and see which questions effectively predict interview outcomes, so we can save ourselves a few interviews, and maybe remove ineffective ones. I do like the idea of using our screening tool as a motivated self-training session, but we do have the 10 percent dropout rate to consider.
posted by pwnguin at 7:46 PM on June 13, 2015 [1 favorite]
This thread is closed to new comments.
In my experience, about half of offers are rejected, with 2-3 candidates for each position interviewed. Once, I estimated ~160 hours effort to hire an engineer, which will easily push $30,000 in actual costs. Of course, most engineers aren't very useful for at least a few months, which adds to the cost of hiring an engineer. In an extreme example for a highly specialized position requiring an uncommon skillset, I know of one instance of a $25,000 bounty to just get an candidate to the interview stage of a position.
I don't know what a "false positive hire" is.
posted by saeculorum at 11:08 AM on June 12, 2015 [3 favorites]