Hacker News new | past | comments | ask | show | jobs | submit login

I find it strange, as software developers, that we discuss the effects on automation all the time, but no one seems to understand that automation has made searching for new employees incredibly inexpensive and efficient compared to the past.

So in technical roles there are a lot of companies that are always hiring. Recruiters work for free until they find you someone and placing job listings online is so inexpensive it barely registers as a cost.

On top of that, most states are "at will" employment, so even if a company does not have any openings, if they find someone better, and cheaper, than another employee that is currently working at the company, they can hire them and then eventually let that other employee go.

For example, I'm not in the market for a new laptop, but if a brand new, top of the line laptop came up for sale at half price, I'd probably buy it.

Its so cheap to look for new talent due to automation, so most companies will shop around in case they get lucky and land a fantastic bargain. Many are not seriously looking to fill an immediate need.

With that said, there are still plenty of companies out there with an immediate need, so please do not get discouraged by that group of companies that are "always hiring".




Interviews take 2-3 man days from a company per candidate you end up doing a full onsight interview. That is on the order of $1000’s per interview, and you generally have to do several per role. Then for high demand roles even the best will take several months to ramp up which costs a large fraction, say 25-50% of their total compensation. So replacing a dev can cost you in total $100k+ at least.

That is not really cheap, I am sure technology has helped reduce the amount of man administrative man hours, and helped fill a pipeline, but you still need to train new hires and actually interview them, which is no cheap.


This happens because high costs that are not easily measurable (such as the productivity hit from interviewing candidates) don't show up on spreadsheets and thus don't result in recriminations from on high. But the relatively low, one-time cost of a new hire shows up immediately on the spreadsheet and that means whoever did the hiring is going to get yelled at for paying too much. Given this incentive structure, it makes sense to drag out the hiring process as long as possible.


I agree with this, but practically no one ever does this math. (Paradoxically, even the people who claim to understand this, and who mention how expensive interviewing is -- even they still fall back into the old patterns)

Interviewing at all companies is always "free", because once your an employee, your time has "no cost", irregardless of what anyone's wages are, and regardless of any opportunity costs for the time.


I agree with your calculations but some managers seem to not have done this math yet.

As result, what we have are poor hiring evaluations, horrible onboarding, etc.


It's exactly those managers that are the problem. They're incompetent and not needed for self-managed agile teams. So they keep rotating developers to safeguard their positions.

Meanwhile those managers are getting the salaries that better be spent on a team of great developers.


I think it is also how budget for hiring seems to be allocated. New funding round, hire a bunch, got budget approved, hire a bunch. Senior management and above seems to not like to give money with time conditions against their expectations.


There's some derived value in the interviewing. It might be to gain insight from an expert in a field. It might be test out interview/hiring practices. For OP, it seems like they were never really intent on hiring someone at the time but put out a lowball offer on the chance it would get picked up.


I wonder at what point would it make sense to start exposing a part of your "software development pipeline" to the external world and then hiring people after they've proven themselves on your pipeline (with compensation). Think lambda school but for your own company.


Open source projects started by or significantly invested in by tech companies must be a huge recruiting tool. I can think of several people in the React community that ended up working at Facebook.


The current data we have about hiring practices seem to disagree with you. So either, this process is not as expensive as you think it is, or companies believe it is less expensive than it actually is.

Either way, the effect on hiring is the same.


searching is easier in terms of internet ads and reviewing candidates. But hiring dev candidates is extremely hard. Virtually everyone has a job who is experienced already, college students get offers somewhere in the early junior year, so you hire them with at most 2 internships. It's very difficult to pull people out of an existing job, it's kind of begging them to consider you. Then for random applicants to a company, you can check their resume, but many people end up failing our interview.

So it's not easy to actually hire. I'm not at amazon, but we do offer reasonable rates - example, new college hires get 130k+stock, benefits at our company.


Are you in the Bay area or Seattle? That's higher rates than what I have seen for new college hires (I have seen in the 40-60k range).


We compete with top companies, although we are a startup. Don't want to say the name. But each year it creeps up. If you want top students who work on backend infra, that's what you pay.

Yes, we are in seattle and/or sf. Don't want to say the name.


$60k seems low to me for any decent size city in the US too. From experience, new grads were offered much more even in late 00s, after recession started.


I made 52k fresh out of college 16 years ago. Seems really low today.


Hmmm. I don't have a ton of data points, but we:

* don't care if you have a CS degree (we have hired plenty of bootcamp grads)

* are 100% remote which is super valuable to some folks

* aren't looking for super specialized skills--we build websites and web applications on some common open source stacks.

* explicitly don't pay top dollar. As the CEO says in the typical hiring conversation, every so often we'll be talking to folks who are also talking to Microsoft and Facebook, and we'll quietly back away--we choose to compete on different axes than dollars.

But maybe we're on the lower end. We definitely pay above that band for folks with 1-2 years of experience.


Fwiw, when I interned in a suburb of Atlanta, as a sophomore at a company no one cares about, I made $25/hr. I believe the people who went there full time started at ~75k.

So I'd consider 60k low for any urban area.


Even at the top end of that range it's less than what a competitive internships annualised compensation would be. Are you really able to hire people at those rates?


>Are you really able to hire people at those rates?

Not the poster,although I've found a lot of people will take 50% pay for fully remote jobs.

I was suprised just how valuable fully remote is to so many people.


> but many people end up failing our interview

If you are having this much trouble hiring, are you 100% sure your interview process is correct?

What is the KPI for your interview process?


That’s another hot topic here in HN. Not claiming is the GP case but interviewing nowadays feels like human Tetris.


I do know we must miss some good candidates. We have been discussing lowering our bar, I'm in favor of changing it. We do the usual 5 or 6 questions. One thing that is interesting about our company is we have a list of x questions (x is less than 20), and we generally ask the same questions, everyone is calibrated on them the same way. Most people who make it through the screening pass the interview loop and get an offer. All our interns got offers the last 2 years. Maybe half the industry people say yes and a smaller percentage of students. For students I think we just don't offer enough.


> We do the usual 5 or 6 questions

But what metric are you determining that those 5 or 6 questions are correct?

For example, have you put together a pool of current workers that are successful at your type of work at different companies and seen how many can pass your interview test? Its pretty easy to find a recruiter that will reach out to their network and you pay 10 employed quality engineers $1,000 that do similar work to attempt the questions. If you find a significant amount of successful engineers cannot do it, then you may want to recalibrate.

I have helped several different companies fix their hiring process, purely by recalibrating their interview test questions to their specific needs. This sometimes actually means, more difficult or simply different types of questions. But very often people just do some Google searches for interviewing at FAANG and repeat those questions without tailoring them to their specific roles.

With that said, in the case of your company, it may very well be the case that the work you do requires particularly high IQ candidates, which is actually what most interview questions at the big tech companies test. In that case, you will always have a difficulty finding candidates as there are only a limited amount of high IQ people in the world. So changing your interview questions may not apply. But when that is the case, you can often save yourself quite a bit of time by asking potential hires to take an IQ online. This will cost your organization almost nothing and will likely screen out the vast majority of people that will not be able to pass those types of interview questions.


I think it’s illegal to hire based on IQ score.


Are you being sarcastic? At least in the US, you aren't allowed to discriminate based on a protected class(race, religion, sex, age, physical/mental handicap) but iq isn't a protected class.


To be clear, while not outright forbidden, the IQ tests are instead called "personality tests" or "aptitude tests" or "coding interviews" because of this landmine legal vaguery (for example, you mentioned mental handicap -- i.e. lower IQ score -- as a protected class).


Sorry, I didn’t mean to deflect the convo towards your company’s hiring practices.

What I was trying to say that lot have been said about hiring metrics and standards around here and elsewhere. Yet, at the end of the day still like arbitrary in both ends and makes the job transaction cost higher than what could possible be.

Anyways, as I read your company’s interview process. Something came to my mind, why not let the candidate decide which questions to pick? I say so because I find the a lot of interviews are so hit or miss but if I was given a bit of control over the process, things could be different (or at least I would like to think this way).


As someone in Canada looking at an offer for much lower than that, I'm not sure what to do to negotiate upwards. I can't find concrete enough numbers that others are making up here for example.


This isn’t right. The cost of finding an employee at my current company (which Is considered an excellent company in the games industry) is near $70k.

The reality is outside of very early resume screening very little of the process is automated. Recruiting spends most of their day contacting people on sites like linked-in trying to find talent as not nearly enough shows up at our door.

My colleagues at FAANG companies tell me similar stories.

The cost of replacing an existing employee is also very high. In most domains can’t just plug that cheaper developer in, they need training and expertise building.

Outside of some very simple domains I don’t think your hypothesis holds up.


And yet, despite all the difficulties, most job ads just look more or less the same- 3 pages of tech requirements, a ton of self praise about the company and 3 lines about what successful applicant would get: bean bag in the corner, Foosball and a spot in an open space office...


I think the talent is intent on showing up. They’re just absolutely atrocious at detecting it even when it is right in front of them.


If it's cheap for them to waste our time with bogus interviews, it's also cheap for us to compile lists of deceitful employers and recruiters. Let's blackball them.

There is actually a Dutch site that tries to do something like that [0], but those recruiters are not even remotely as bad as these stories. They may be annoying parasites, but at least when they mention a number, it's fairly reliable.

[0] https://blacklist-recruiters.nl/archive/


"I find it strange, as software developers, that we discuss the effects on automation all the time, but no one seems to understand that automation has made searching for new employees incredibly inexpensive and efficient compared to the past."

Yes, just like Tinder and other assorted dating apps and sites have made it incredibly easy and efficient to search for a mate, but relatively few people end up in a relationship because of it. The act of swiping endlessly appeals to people, apparently. I vaguely remember hearing that once you are given 2 or more options, it induces strong paralysis, whereas if you have one option you tend to just jump at it and move forward with your life. I imagine the dynamics of dating and hiring are quite parallel.



> once you are given 2 or more options

You mean 3 or more? Because saying 2 or more just means you have options and therefore any options induces paralysis


only a programmer could produce such a pedantic comment such as this.


> Anyway... Went in for 4-5 interviews. Took me about 10 hours in total to interview with them.

The experience of the poster you are responding to doesn't seem to align with the situation you are describing. Sure posting an online job ad is pretty cheap. The rest of the hiring process, not so much.


Believe it or not, those interviews usually aren't that expensive. You probably have a mix of managers, which spend all their time in meetings anyway, and engineers, which are paid salary so they typically end up putting in a couple extra hours when they fall behind due to interviewing.

Sure it isn't free, but remember, this person had made it all the way to the end of the process. The determination was made that they represented such a large gain, for the low salary they were paying, that it was worth the risk of that small cost to get an employee at a 30% reduction in salary.

If you are going to save 30% of $140k, that means you will save $42k. It would take a lot of interviews to make up that gap.


>so they typically end up putting in a couple extra hours when they fall behind due to interviewing

Mind letting me know where you work so I can avoid it? Interviewing is part of the job and needs to be accounted for. You don't just make your employees do overtime because you don't want to plan out your companies work well


Yeah... no. Most engineers and programmers don't interview for free. They do it on company time.

If their other work falls behind because they spend too much time interviewing, well that is the company's problem. I have never worked with anyone that would stay late because they interviewed someone that day. That is ridiculous.

And as for staying late "to get the work done", what job do you work in where the work is ever "done"?


If there are 10 hours of interviews, and each one is attended by an average of 1.5 current employees, who make, on average, $100K, then the labor cost (not including opportunity cost) of that round of interviews with that candidate is on the order of a few hundred to a thousand dollars. So you might have a point there.


Searching for new employees is incredibly inexpensive == I now know you don't do any interviewing or hiring.

Eg when we post a senior eng job on eg glassdoor, stackoverflow, or linkedin, we are deluged with resumes. Just the ad easily costs $250. Then we get applications who ignore that we don't sponsor visas, clearly don't have the skills required (6 year SAP programmer who wants to be a senior fullstack eng? No thanks), or are junior eng trying to make a huge jump. 1.5 years experience does not make you a senior SE who can be trusted to design and land large features with minimal oversight.

The first problem is an eng manager reviews all those resumes. It easily requires 10 minutes per, including writing nice rejection letters. And that's just for inbound. Outbound can take 30+ minutes per.

If someone passes the resume filter, we do a phone screen with one of our senior eng shadowed by a junior. A 45 minute call plus prep plus decision afterwards costs me 2 hours of eng time.

If you come in, we do about a 3.5 hour interview across 6 people. We cover travel expenses and lunch if an interview runs over lunchtime.

The whole thing easy costs us $1.5k, plus tons of eng time. And that's if we source. If a recruiter sources, we're looking at 20%.


>6 year SAP programmer who wants to be a senior fullstack eng? No thanks.

Which part of senior full stack are they missing? Front end, back end , API?

https://cloudplatform.sap.com/capabilities/technical-asset-i...

>1.5 years experience does not make you a senior SE who can be trusted to design and land large features with minimal oversight.

Yes and 25 years experience doesn't make them one either, especially if it is the same 25 years.

You could find a 2.5 jr with the skills required.

I mean if they started their own start up they could be designing and landing features with no oversight within 18 months.

>The first problem is an eng manager reviews all those resumes.

Yes it's better to have actually engineers do the initial review. They understand the job, will see through the BS and scale better. The more people you hire, the more resumes you can review.


$1.5k is a significant expense for your organization? A $250 ad listing is a lot for your organization?

I'm sorry, but this sounds very strange from someone claiming to know about interviewing and hiring.

You need an engineering manager to filter out SAP programmers who want to be a senior full stack engineer? You need an engineering manager to figure out someone only has 1.5 years of experience?

How many people are you interviewing? How many positions are you filling? Seems like you would realistically need about 5-10 phone screens plus 1 or 2 in person interviews to fill a position. 3 if your phone screens were really off the mark.

I'm actually curious who hired you. Because this sounds like dismal operations management. Seriously, you seem to be a great example of how much a bad hire can cost a company.


Reading the original post it seems to say 1.5K plus "tons of eng time", I'd imagine that it's the opportunity cost of the engineering time that ends up being by far the most expensive part in this and that the 1.5k figure was just the other fixed costs.

"Seems like you would realistically need about 5-10 phone screens plus 1 or 2 in person interviews to fill a position. 3 if your phone screens were really off the mark." Just curious what your thought process is to arrive at these numbers?


This should have been obvious, but buttresses my thesis: we do more than one ad. Run rate order of $140k for 2019. obviously $2k is not a significant expense, but that's per position per interview. So we pay that well more than once.

Also in this thread: piles of people complaining hr runs hiring. We have an eng manager run hiring for our eng. You: bring back hr to review resumes!.

We're hiring on the order of 15 eng over the next calendar year. Ask anyone who has actually done this (say, for example, the folks at yc) hiring is incredibly expensive. I saw an interview where an Asana founder said he spent perhaps 20% - I can't recall, but significant percentage -- of his time hiring. That's super senior executive time focused on getting employees in.


I wonder if automation actually distorts the recruitment market? As the initial cost of recruitment goes down, thanks to automation, more recruiters gets involved because it's cheaper to enter the market.

As more recruiters offers their services to employers the employers starts to feel like there are more potential hires than there actually are.

As a thought experiment we could assume that there used to be two recruiting firms for each employer but then costs went down and now there are ten. This gave the illusion that there are five times as many potential hires even if everyone should take a step back and realize that the recruiters are still all trawling in the same pool in their search for talent.


I don't totally buy this argument. It's not cheap to hire. They might have increased the number of candidates in the pipeline sure, but I don't think the turnaround time to hire is any less. Also, there's the paradox of choice. The more candidates you have, the harder it is to hire. You actually need to apply filters to reduce the overall quantity you will pick from.


There's also cost of switching itself which could be pretty steep. Pay someone to learn his role being essentially useless for the first month or so, and having several other people distract to show him/her the way around.


You've clearly never worked on the hiring side of a company. I mean that literally and not snidely: you're completely unaware of the costs and complications and risks that go into hiring from the company's side.


those are some good points, I been wondering if they are doing things like that myself.

I also wonder how much of it is also #1 data gathering, #2 advertisement, and #3 slightly counter to what you are saying - bad tools/people using said tools.

for point one, I can learn a lot on what microsoft is doing by pulling resumes from microsoft workers, even more if I interview them. you can quickly glean direction, head count, tech stack, etc.

for point two, tech companies especially have to remain relevant in the news. Imagine if you post positions for great ML shiny positions, that gets seen by tens of thousands of people for 0 cost. "Did you see that great posting at company X? wow they have an impressive Y team"

point 3, I have seen recruiters get thousands of resumes for a position they posted on the web, really want to hire someone, but instead send people through the hiring loops because of direct contacts. Its like they search for skills, the resumes on the website dont get searched well so they assume there are no matches. its baffling to me this happens given the tools you mentioned




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: