Documents like these shouldn't be automatically processed, they should be reviewed by humans. Reducing someone's life history to a list of educational institutions and employers feels robotic even for a software developer's mindset.
I understand that there are real life problems because companies do use automated processing on applications, but that kind of behaviour shouldn't be encouraged.
>Documents like these shouldn't be automatically processed, they should be reviewed by humans.
That's very nice to think about but in the real world if people don't match some exact keywords they get thrown away. HR is looking for "node.js" but you wrote it like "nodejs"? Though luck pal, bye. HR wants a "computer scientist" but you have a degree in "computer science"? Same, automatically discarded.
If anything, a standard format would at least allow people to classify themselves/others correctly and without ambiguity.
HR is looking for "node.js" but you wrote it like "nodejs"? Though luck pal, bye. HR wants a "computer scientist" but you have a degree in "computer science"? Same, automatically discarded.
Fixing that would require not only a universal data format but also a universal taxonomy of terms to describe skills. That would need everyone to agree on how to describe what they do. Easy enough when you're describing programming languages, but effectively impossible for any softer skills. This is precisely why you need the human aspect.
The way to fix the issue you describe is to educate HR people, not to try to apply technology to a non-tech problem.
I'm pretty sure my Resume says "crazy outside-the-box madman" on it somewhere near the top. It pretty much guarantees that I work somewhere with a sense of humor while rejected from places that take themselves too seriously.
Resumes aren't just a list of skills and work history (education isn't even on my resume anymore), it's also a chance to let a small amount of your personality show through so you stand out against the pile.
Of course, if you're submitting to a big corp, you'll be fighting against AI reading your resume first, so, write for that if that's how you roll. I usually only apply to smaller companies that apparently end up as bigger companies by the time I move on.
My LinkedIn recruiter spam dropped substantially when I added "literally a wizard" to the skills section (and removed everything remotely tech-telated).
Fixing that would require not only a universal data format but also a universal taxonomy of terms
A regular HR can only do CTRL-F CTRL-V efficiently in their PDF viewer, when automated system would likely test for /\bnode(?:\.?js)?\b/i, because it's a result of a development cycle, which usually includes some field analysis. HRs would LOVE to select something like "where has($node) and years($redux) >= 2 order by years($php) desc". They are not idiots, they just have piles of data and nonsensical tools. Of course they optimize for nonsense.
A human could search for different variations too, but they are average in average and have no the expertise to install a regex-able viewer. It's stupid little things like this that prevent them from doing their job better, not some systemic issue of their own.
I can see why. Of course a role that requires CTRL-F level of skills attracts low-skill-required resources to a "helper" subrole. But if you replace it with automation, true skills will remain and thrive. Not that I like them on emotional level, their job is to hire you as cheap as possible and demand as much as they can. But hey it's a market. Learn to negotiate and to plan ahead financially for search periods to avoid premature decision pitfalls.
Maybe I've not applied and worked with big enough companies, but I've never seen an HR person be helpful or in any way been aligned with my interests. This could be forgiven as they may well serve interests which are opposite to mine at times, but there are numerous times where HR representatives just (from what I can gather) pick suboptimal strategies to try and convince me to do something, be it remain or take up a job offer or anything else, i.e. I've yet to see an HR person successfully manage a human resource in an optimal way. And by that I mean that statistically the outcomes would be no different if it was a random person with normal communication skills doing the job of HR or if it was someone with training and experience.
> The way to fix the issue you describe is to educate HR people, not to try to apply technology to a non-tech problem.
I wish it was easier to do. At my first job after university (not my first job altogether) we initially had great rapport between "DCOps" and HR. To be honest, our group was rather... irreverent, which made it even more interesting that we had such great rapport (we had all the stereotypes going for us, including female-only HR team and basement-dwelling male-only sysadmins).
But thanks to great cooperation, HR would do basic filtering on things they actually checked with us about, then forwarded us the CVs unmodified to give our own opinion before deciding whether to go with it or not.
Unfortunately some time later head of HR ended up let go due to some conflict with C-level I believe, and things slowly reverted to the mean :(
That's very nice to think about but in the real world if people don't match some exact keywords they get thrown away.
This isn't the "real world" you're referring to here -- it's the fin-de-big-tech bubble world where everyone and their dog is not only sold, smack-down drunk on the idea that Algorithms and Data are the Solution to Your Problems. They aren't of course -- it's just a giant hornswoggle. This "world" needs to end, and it needs to end soon.
My belief is that it will. But until it does, companies that run crappy ATSs (like there's any other kind) deserve the buzzword-gurgling, keyword-dropping candidate pool get. And developers who get "rejected" by these companies should be grateful for the sublime gift of this rejection -- and for the opportunity to laser focus on companies and teams that use their heads to hire, rather than a fleet of bots.
Snark aside - hiring managers that actually read resumes (and yes, actually at least skim each and every one -- really it ain't that hard) are golden to work for (other factors being equal). Really, you don't want to waste your time with companies that have drunken the ATS kool-aid. Really you don't.
That said, however:
HR is looking for "node.js" but you wrote it like "nodejs"? Though luck pal, bye.
It's "Node.js" with a dot. If you spell it "nodejs", that's a serious red flag and you shouldn't be surprised if it gets you automatically discarded -- even by a human reviewer.
That's what resumes used to be, after all -- a kind of a take-home test where you really do have enough time (and perfect knowledge) to get everything 100 percent right. And also a test of your awareness of the fact that, yes, in critical business communications at least -- this level of correctness does matter, and it matters a lot.
> It's "Node.js" with a dot. If you spell it "nodejs", that's a serious red flag and you shouldn't be surprised if it gets you automatically discarded -- even by a human reviewer.
Yikes. “It’s not JavaScript it’s Ecmascript! What a RED FLAG.”
The difference is that JavaScript is the name of an actual language. Where nodejs isn't the actual name of the project. The Node.js project advertises itself as a "JavaScript runtime." Both JavaScript and EMCAScript are acceptable.
I’d say they are. Who cares if I write “Nodejs” or “Node.js”? The reader knows exactly what I meant, and that’s what matters. Anything else is just needless pedantry.
It's not about communication, here. You want to be absolutely perfect in your CV, to prove that you've spent time and effort making it perfect. It serves no other purpose but to show your determination. If communication was the point, a long rant in which you write "yeah I worked with node.js, it was cool lol" would be OK as well, after all it gets the point across doesn't it?
The things is -- they ding you for a whole range of other things, all the time. Like not working for the right kind of company in your last job, failing their bullshit tests, or heaven forbid, having a "gap" in your resume. As if having a life (or a family member that needs taking care of) is to be seen as thumbing your nose at their sacred cause.
At least fixing stupid mistakes in your resume is well within your control, and (unlike cramming for their tests) doesn't take hours and hours of your time.
That's fine. I see it like this: If there is a stupid mistake, it shows that I am a fallible human after all. (I tend to think that I don't have stupid mistakes in my CV)
But my CV also shows part of my character, mainly the fun part. If that is deemed unprofessional by a company, I immediately know they are no good fit for me.
The thing is: As with every relationship, they are two way streets. I am not a beggar as a potential employee, neither am I king. We see eye to eye or not at all.
I'm an OCD-ish perfectionist (especially in writing) if the world has ever seen one, to the point where I didn't stark working on my bachelor's or my master's thesis before I had exactly the markdown + latex setup + custom typography that I wanted. So I would definitely internally cringe if I'd see typos like this.
But to reject a candidate just because they're not the same kind of pedant that I am would be quite unreasonable IMHO. As long as the whole thing isn't littered with typos or formatting issues, that's just something I wouldn't pay too much attention to. Not everyone needs to be the person to dot every i and cross every t.
Also, I really want to see the HR department that can confidently reject or even mark down a candidate when most of the writing I've seen from most people in any company I've ever been (including from HR departments, PMs or executives) has been "meh" at best, and full of errors or at least awkward language more often than not. For better or worse, writing skills are not necessarily something we tend to select for in the industry.
It really does require an eye for detail (in places where it matters) to a degree that comes epsilon close to being "pedantic" in the annoying sense -- without quite going over that line.
> That's what resumes used to be, after all -- a kind of a take-home test where you really do have enough time (and perfect knowledge) to get everything 100 percent right.
The only thing this filters for is people who have been looking for a job for a long time and/or are people focused on presentation instead of content, good luck with that.
We’ve hired incredible engineers across that have had typos and major resume mistakes many times and they had no communications problems on the job.
A resume is a background/interests check for the hiring manager. Giving them a spell check is about as useful as measuring the ratio of times the letter ‘a’ appears relative to ‘e’.
> critical business communications at least -- this level of correctness does matter, and it matters a lot.
This sounds like someone who is unable to contextualize when that would be relevant. Unless you’re writing user documentation there are zero times in an engineer’s day-to-day communications where writing nodejs instead of “Node.js” would have any impact at all.
Giving them a spell check is about as useful as measuring the ratio of times the letter ‘a’ appears relative to ‘e’.
I'll leave the matter of whether one should be expect resumes to be at least reasonably correct (in terms of spelling and grammar) or not - or what the threshold for "reasonably" should be on this front - to the side for now. It's obviously a question of taste, to a certain degree.
But sorry - what you're saying about the ratio of 'a' to 'e' makes no sense at all.
And on this front:
Unless you’re writing user documentation
I don't know what planet you live on -- but on my planet, good engineers write "user documentation" (in the sense of READMEs, docstrings, etc) and participate in mailing lists, do conference presentations, etc, all the time. And yes, while nobody is 100 percent perfect, consistently making a large number of gaffes (or more to the point: conveying an attitude that these things just don't matter; or that you're so hot that they shouldn't matter for you) in these contexts -- particularly when there there is plenty of time available to edit and review -- is definitely seen as a red flag, both in and outside the company.
These days I think I work at the best team I ever worked on and it seems I'm in the top two in spelling, it is almost as if the rest if the team is mildly dyslexic or doesn't care.
But they are nice to work with, smart and they make good code.
I'll take them 49 times out of 50 over some perfect in code and writing but nasty in person ex colleagues.
Even if I really really really need a specialist on hight performance <whatever it was, I can't add a dead giveaway here> I'd still try to look elsewhere first.
Edit:
... and those two ex colleagues? Their code was good but it wasn't perfect either.
Am 100 percent with you there - nice and basically decent to work with always trumps the willful disregard for these basic qualities. No matter how skilled they are, or their accomplishments.
> But sorry - what you're saying about the ratio of 'a' to 'e' makes no sense at all.
Sorry I confused you. Let me be clear, it’s not relevant at all to job performance. I would take someone with terrible spelling and punctuation who actually writes documentation over someone who writes half as much because they are obsessing over details like that.
> in the sense of READMEs, docstrings, et
These are generally for other developers. Developers do not get confused when they see “NodeJs” instead of “Node.js”. Do you know what point you were trying to make?
> is definitely seen as a red flag, both in and outside the company.
> For any company you'd want to work for, that is.
Not sure what kind of company you’re talking about, but it’s definitely not a tech company. Maybe Wall Street or something a little slower paced?
It's not about communication, or spelling, or grammar, it's about attention to detail.
I've seen my fair share of resumes and hired my fair share of people, and I've come to the conclusion that sloppy, error-ridden, CVs are a sign of a lack of attention to detail.
Here's a candidate with all the time in the world, and huge motivation, and they still can't complete this task correctly. It's not a good look.
I'm not saying a single error I'd a deal-breaker, but it puts the applicant on the back foot. No doubt many of my rejects are quality engineers, but engineering, and software development are all about details.
> Here's a candidate with all the time in the world, and huge motivation,
You are woefully out of touch with the job market. And I’m saying this as someone who worked as a hiring manager at one of the most applied to tech companies in the world.
The motivation might not be "huge" at all, considering you might not be expecting any response, might have done 10 of these in the last hour, pay may not even be acceptable, etc.
A job hunt typically takes weeks or months of calendar time, and dozens of hours of clock time. So if a candidate has good management of his time and effort, he can budget one of those hours to improve the attention to detail of his CV.
>It's "Node.js" with a dot. If you spell it "nodejs", that's a serious red flag and you shouldn't be surprised if it gets you automatically discarded -- even by a human reviewer.
Oh, really!? I'll be very glad NOT to work at a company that throws aways candidates because of such triviliaties.
Even back in the era when reading resumes and hiring was a manual human process, the standard advice was that resumes should be perfect. Nicely formatted, with no spelling or grammatical errors. You have ample time to think about the content, to proofread it yourself, and to have someone else proofread it. It's the first impression the potential employer has of you, no reason to not make an effort.
I'm not sure that's true at all. Coming out of school pre-Web when real information was much harder to come by (even job postings), I sprayed and prayed a lot of resumes to HR departments. (And my first engineering job came through one of those.)
Perfectionism and pedantism are good qualities some teams are after, it is unsurprising that some people would reject a candidate who did not bother “properly” writing the name of the technology they use.
nodejs came out on 2009, I've used it non-stop since then and probably built close to 100 things with it. I am on the author's list of both node and v8.
I write it as nodejs because that's how I got used to it on the mailing lists and several places that I frequent.
If a "team" decides not to hire me because of the latter, then ... thanks, I guess?
I guess I'll backpedal at this point and acknowledge that that "nodejs" may in fact be not only a correct and accepted spelling -- but in fact a hip "in-the-know" spelling -- within a certain community.
But at this point we're getting into the territory of what's seen as a positive in some niche communities may not be seen as a positive signal outside that community. And in business communications, what masters most is knowing who your audience is, and how your message is received.
Then you being outside the community, have done nothing but filter out the best candidates by focusing on the wrong details under the banner of “attention to detail”
It’s ego.
What is the résumé’s job? To convey experience and work in a positive light. A small typo that may incidentally be the correct spelling shouldn’t derail that if the goal is to hire the most talented people.
To my defense, I was going by my Bayesian prior for certain language names for which one could be 100 percent sure there was no alternate accepted spelling or capitalization.
So if one put JAVA or PERL on their resume, that would generate a definite "pass" signal.
Agree or disagree, this is great feedback for job seekers. Spelling is often used as a filter.
What are some other signals that act as red flags for a potential applicant in your hiring flow?
Well another major turn-off would be: a "primary skills" section that mentions 30 different random applications / protocols that one might have been incidentally exposed to for a week or 2 at some job ... 15 years ago. No one can be expert-level in all of these things. Just tell us the top 3-6 that really matter and which define you.
Or even if it's just 1 or 2. It's infinitely better to hire someone who is in fact really solid at what they say they know, than someone who tries to spam you with every random keyword they've been exposed to in the hope that you won't drill down and ask them any hard questions about most of them.
That, and skill listings that don't pass the "apples and orange" test. For example:
Something tells me they could easily focus their attention on something not necessarily related to the most critical parts of the project. They could be very pedantic (and often are) about some irrelevant thing.
I have a lot of merge requests in some pretty big projects correcting SASS to Sass in the code and documentation. I'm surprised no one ever seems to question their notion of spelling proper names when it's a quick web search away.
The former I think you’re spot-on. If you’re serious about hiring, some qualified to do so needs to be reading resumes.
The latter is complicated. I tend to agree that the industry has over-pivoted to “dynamic programming whiteboard puzzle” as the LSAT of high-paying software work. But while most seem to agree that leetcode interviews suck in tons of ways, something that is unambiguously superior seems to be an open problem.
And some algorithms and data structure testing is important for hard software work. Even choosing the right library requires some basics. Is every SWE at Amazon implementing CS papers every month? No, clearly not. But unless machine resources are free in your domain, brute-force doesn’t always work.
FANG people do in fact get elitist about their CMU educations. But there’s another crowd that throws the baby out with the bath water: “I solve business problems, not reverse linked lists.” Well, what kind of business problems don’t involve some computation?
It’s a balance that our industry seems to still be trying to find.
So, when I see the ads with XML misspelled XLM like XLM and XSLT experience wanted - or similar stuff that happens every now and then - I should spin up a new CV and replace instances of XML with XLM?
on edit: before people come in with not wanting to work for a company that would make such a stupid mistake, first off I believe mistakes sometimes happen, and second Henry Thompson leader of the XML Schema working group, had the subtitle of his LinkedIn page saying something like he was an expert in XLM for at least a year (I should really have taken a screen shot) so I mean I think it would just as silly to not work someplace because they made a mistake as it would for them to screen you out for making a mistake.
These are true:
- hiring managers that actually read resumes are golden to work for
- a test of your awareness of the fact that, yes, in critical business communications at least -- this level of correctness does matter, and it matters a lot.
And for people who want to work in startup land, it’s an ideal.
Once a company gets past 40 or so people, though, ATS adoption increases. And 150+ employees without an ATS are rare. It’s just that the volume of candidates and interview processes that need to be managed require there be some software to manage the workflow.
> It's "Node.js" with a dot. If you spell it "nodejs", that's a serious red flag and you shouldn't be surprised if it gets you automatically discarded -- even by a human reviewer.
Best example of someone I wouldn't want to work for ever
The question is does that really bring something good, either way, or is it just a whim. Some HR check the shape of your ears, and if it doesn't match some patterns, they skip.
perfect filter, both ways
Two ways of three. Some innocent person may accidentally put Node.js in their resume and get hired.
Google has 411m hits for Node.js and 66m for nodejs. I'd guess somewhere in the 66 million people calling it nodejs, we could find a real diamond in the rough if we heard them out at the next interview!
That's exactly what I mean -- people believe in Big Data so much these days that when something appears in a sufficient portion of search results, it must have actual validity.
Isn't that the same as natural language? "Bone Apple Tea" gets corrected both because it's wrong, and because many people know so. "Begging the Question" is used incorrectly _so frequently_ that even dictionaries have come to accept the incorrect usage as an alternative. It's now correct-ish just because it's popular.
So if Node.js has that many more hits than nodejs, it's not surprising that some people see it as canonical.
Hearing the context of this conversation, though, I can see that "nodejs" in your resume is a signal to senior peers within that community--others who are "in the know". That may be a useful filter.
As an aside, I never thought "begging the question" is used incorrectly because it is easy to tell whether or not someone is referring to the fallacy based on the context. If the meaning can be understood, then that's what matters.
Absolutely. We typically receive hundreds of resumes for every position. The easiest way to winnow them down is to throw out any with spelling or grammar errors. It is astonishing how high a percentage include these errors. There is absolutely no excuse to have spelling and grammar errors in your resume if you are serious about trying to find work.
We rarely receive resumes from people who are not native speakers. It is the native speaker who make the worst spelling and grammar errors. It is so strange in this era when every computer has a spell checker and grammar checker. They aren’t perfect, but most of the errors are so egregious that they clearly aren’t even trying.
It’s a document sent to tons of people. Revised by tons of hiring managers. Can’t you spend 5 minute looking at a dictionary? If you cant be bothered to spend 5 minutes researching this for an important document, what makes me think that you will spend 5 minutes trying to be good at your job. Lack of care on your resume is lack of care on lots of other things.
Companies don't seem to treat it like an important document. I can't tell you how many times an interviewer seemed to be reading my resume for the very first time during the interview itself...
As the interviewer, in several places I've worked, I'm usually handed the resume 10 minutes before the interview along with a couple of questions my manager wants me to ask. By that point, it's gone through HR, managers, etc. You're getting hired if you can get past me and whomever I'm interviewing with.
There was even one time they forgot to give me the resume... that was probably the funniest interview ever. They got the job too.
On the flip side, I do technical interviews (I am new to it though), and I spend about half an hour preparing for an interview including going through the CV that made it to me (and making notes/preparing questions) and researching the candidate based on what I see in it (e.g following links, etc)
I do the same thing. The CV is typically made available to me a couple of days before the interview, and I usually take a look at it the day before. Obviously I can't really know how carefully it gets scanned before they decide to arrange the technical interview (which is where I step in).
If you cant be bothered to spend 5 minutes researching this for an important document, what makes me think that you will spend 5 minutes trying to be good at your job
Bullshit. There are many things person may care about and many things they may not. If this position is unrelated to professional writing, it's just irrelevant. Are you questioning their abstract ability to "care" and trying to translate spelling care to job care? Bad news, it doesn't translate (aside from the fact that ideal perfectionists can't get shit done ever; your best candidate still figures out the ideal form of his resume and is too anxious to click "publish").
Native speakers are the worst at pointing typos in my experience because they are so sure of themselves.
Literally every time I have submitted an (academic) paper at least one US reviewer had to point out I made several spelling mistakes when I actually used the British spelling (consistently) instead of the US spelling.
Sorry my man but it is called "English", I think I am allowed to spell things like they do in England.
Some people aren't native speakers. Do they really need to know how to spell perfectly to work for you?
Perfectly, no. But in formal communications, they need to show that they're at least making an effort to get things reasonably correct. And if it matters enough, you can always find a native-speaking friend to review your resume or cover letter for you.
(Source: been there, done that, applying for jobs in other countries).
I think it's perfectly reasonable to want to filter out a candidate that can't be bothered to spent 3 minutes and run their resume through a spell checker. It tells you something important about the candidate.
Of course, it is also reasonable to be forgiving on resume errors. It depends on the culture and team you're trying to build.
Speaking of red squiggly lines in Word/MS products. Is it just me or has the grammar check gotten incredibly worse the past 2 years? It keeps complaining about odd grammar forms and suggesting I replace perfectly valid word-structures with shorter ones? The one I encounter most frequently is the suggestion to replace "in order to" with just "to", which is incredibly frustrating as they mean two different things entirely.
All of the lines below have red squiggly lines under them in my browser's text area:
- Qt Quick
- Datalog
- Datomic
- TensorFlow
> I've noticed that ESL speakers make different kinds of mistakes than sloppy native speakers do.
I've never seen a non-native speaker writing "would of" or using "they're" instead of "their". On the other hand I can't count how many times I've skipped an article or used a wrong one somewhere (since my native language doesn't use them at all).
Yup. My wife learned her first word of English at 43, I've had plenty of time to see how one's mother tongue has a permanent effect on language. My interpretation of what I see:
She has internal word concepts. When the English words fit cleanly into her existing concepts it's easy. When two English words map into one word concept it's considerably harder, it took her a long time to correctly separate turn on from open (and turn off from close), but at least any given thing always goes the same way. In situations where that doesn't even work she will make frequent errors even to this day--do not trust her use of gender words.
Its not just big tech that does this. HR is also incompetent in academia and has similar practices at many institutions
Edit: oh, and also US state governments. there's a reason they're known for being incompetent, people job hop within months and get keywords on their resumes and nothing ever gets done
Well said. It'll get to a human eventually, but the first step is pretty robotic anyway human or not. Let's dig up the legendary Facebook recruiter specifically looking for Unix experience.
>> It's "Node.js" with a dot. If you spell it "nodejs", that's a serious red flag and you shouldn't be surprised if it gets you automatically discarded -- even by a human reviewer.
I kinda agree with you -- half the job as a programmer is being really tedious and paying attention to written text on a screen. If you can't reasonably review your resume and use spell check, it is not a great look.
I review every single resume and applicant I hire. My experience tells me talent acquisition professionals don’t really know how to match and will often suggest poor matches and not bring forward odd fits with interesting backgrounds.
I get hundreds posting into some of the roles I’ve hired for and I’ve got to look at every single one in order to find a possible match. It generally takes me about 30 seconds to a minute to assess a resume. A kernel of interest stands out quickly or a smoke screen of buzz words tells me it’s the wrong candidate.
Algorithms aren’t a substitute for an active hiring manager being really interested in finding the right candidate to join their team.
> HR is looking for "node.js" but you wrote it like "nodejs"? Though luck pal, bye. HR wants a "computer scientist" but you have a degree in "computer science"? Same, automatically discarded.
Wouldn't those issues be more likely to occur as a result of automated processing? I imagine that most humans who deal with technical hiring would guess that "node.js" and "nodejs" are probably the same thing.
> That's very nice to think about but in the real world if people don't match some exact keywords they get thrown away. HR is looking for "node.js" but you wrote it like "nodejs"? Though luck pal, bye. HR wants a "computer scientist" but you have a degree in "computer science"? Same, automatically discarded
I'm sure this happened somewhere at some point in time, but IMO it's becoming a myth blown out of proportion.
I've never actually worked at a tech company that had any sort of automatic resume filtering software. As a hiring manager, we didn't even have HR pre-screening resumes for us. If we did have recruiters, they were competent and incentivized to find good candidates (e.g. wouldn't be unnecessarily rejecting good candidates because they were missing exact spellings of keywords)
It would seem ludicrous in 2022 and this hiring crunch for a tech company (one that you'd actually want to work for) to arbitrarily filter out resumes based on incomplete keyword searching.
I co-founded an ATS company and I can say that we were never asked for, nor provided, any functionality to automatically screen candidates out based on keyword filters (though candidate answers to actual questions like "Do you 5 years experience with node.js?" are often used to automatically screen out).
That said, I have heard this often enough that some people must do it. I can't imagine why though, as it takes only a few seconds for a human to scan a resume. Even at 10 seconds per resume (a conservative estimate), that means the human could screen 360 resumes per hour.
Where do these resumes in the ATS come from? Sourcing tends to use searches or keywords, which “filter” by not selecting in the first place. Applicant trackers tend to be a few steps downstream of that, so wouldn’t need to filter.
Based on experience at firms with on the order of 1:1 applicants to hires up to 10,000:1 applicants to hires, I’d agree most such tools are not present at most companies, as they tend to be used from ratios of 100:1 and up.
On the other hand, while most companies might not have or use such tools, the more jobs a company has, the more likely they are to have some form of selection or filtering in the pipeline.
I’ve interviewed the CEOs of many of the largest ATSs for Ladders News, and while there is no automated system, the tools do exist for a company to screen in this basis.
But none of the ATSs really collect great data on how many of their customers use them in their core processes. From observing actual, heavy duty recruiter and HR usage at Ladders, I suspect it is not very many.
These are people people after all, and they are not great at writing REGEX style queries to find only the precisely correct resumes they are looking for. You also need to remember that they are frequently working on novel searches each week, and wouldn’t have the confidence or the capability to craft such searches.
I'm not sure if I'd mind that. If HR is made up of robots that stupid, that sounds like a terrible company to work for.
A standard format wouldn't solve the problems you described. There's no comprehensive list of frameworks and programming languages, let alone for types of experience outside computer science. The data would be structured, but HR would still throw out the bits of data that need to be filled in manually.
It varies. There is a reason there is some level dislike for HR cohorts. Keyword bingo is one of those things.
Minor anecdote incoming. I applied to a company A. It was a fit. I think I used the right keywords. Zero response. I talked with my buddy there. She talked with hiring manager. I got hired by company A. I later found out HR was unhappy, because it undermined their 'process' - such as it is.
There just has to be a better way to do this when compared to things as they are now.
Going direct to a hiring manager is always the best route to the job.
Any standardized process will lack fidelity and excellence in specific cases. Or maybe even in all cases, when the company has made the trade off for cost or speed over quality.
But that’s what big company life is like, and that’s why they have standardized processes. Because even with the errors, in aggregate, it works out better for the business.
Do you really care that you made HR unhappy though? Seems like you shouldn't, the company/team management is probably happy to pursue references over the funnel of random candidates, and the unhappy HR staffer is better off learning from the experience.
Not really is the truthful answer. The job in question had my best boss so far, so upsetting HR drone was a necessary evil in retrospect. I just dislike making people upset if it can be avoided. In my line of work, I do not get to make a ton of friends along the way. And who knows where that drone might end up in 10 years.
Do you really care that you made HR unhappy though?
Fundamentally it does kind of suck to have to put up people in your environment who are "unhappy" with you, or the fact that you snuck in there -- not because you don't belong there, but because of their own basic ignorance as to how things actually work.
Not that you can't grit your teeth and live with it - it's a job after all, and that's what you get paid for. But still, it does kind of suck.
I believe I understand your point – therefore I would like to take this opportunity to channel something that 'patio11 wrote on Twitter many years ago:
“Do not send in a job application before you have an actual human
being who has expressed enthusiasm for reading or forwarding it.”[0]
I believe this is sound advice, and the pitfalls you mentioned was probably the reasons that he gave this advice.
That's a good advice but probably for the top X percent in their industry who have a luxury to choose their employer - not everyone is in this position.
It's not about picking your employer so much as it is about getting past the outer layer and in touch with an interested human who can champion you through the rest of the process. In the absence of a pre-existing relationship, a good recruiter is invaluable for this.
Recruiters can definitely be a good way. That’s how I landed, like, nearly every job I have ever been offered.
Networking and reaching out to people at interesting companies is another way.
You will probably still have to write a cover letter and send in your résumé. But if you have someone on the inside who will mention to the hiring manager that, ”hey, you should check out Jane’s application” it will probably make a great deal of difference compared to just applying and hoping for the best.
And please folks, remember that FAANG is not the only game in town. There are lots and lots and lots of other companies out there, and many of them also pay well since they need to compete with FAANG to at least some extent.
I remember the HR chief of [famous company] complaining to me that she'd already been ordered to offer me a job, and needed a resume to check her bureaucratic boxen.
I have to agree, I have a lot of respect for McKenzie but this advice seems unwise. I've landed four jobs (including my longest-lasting gig to date) by sending in resumes to job listings without knowing anyone internally.
This is really not an issue. If the company as such a dumb HR department on automatic mode, I’m glad I’ll never be hired there.
But to be honest, I’ve never seen so stupid HR people. Like a lot of people, they have goals to reach and they’ll happily accept an interview from any CV as long as there are computer-related keywords and roughly the required experience.
HRs are one of the most important department when it comes to a company potential shitiness. So I’m glad they have the power to kick you from the hiring process for stupid reasons.
> If anything, a standard format would at least allow people to classify themselves/others correctly and without ambiguity.
Except it doesn't, because a lot of things in life doesn't fit in that boxes that well.
You might not have experience in node.js, but in JS and some of the node.js-alike server side VMs.
Even just a "degree in computer science" can mean many things. Furthermore you might have something which for the position is "equivalent", like a "master of science in cognitive systems with a specialization on AI" which contained a lot of CS courses, and maybe also some pr-axis experience. Or you just started working as a programming free lancer with 16. Or ...
Furthermore a resume is quite individualistic, beyond just the raw content it contains. How that content is represented can sometimes give you hints, about what kind of questions you want to ask during an interview.
I have seen a bunch of standardized forms for resumes some companies opted into, they _always_ caused endless problems.
This is why applying for jobs using a resume is a losing battle. Make connections, get introductions, get recommendations. Plenty of ways to explore new jobs without being resume 123 out of 500 who applied. No telling what arbitrary filtering criteria will knock you out of the competition. Absolutely no feedback on why you weren't considered.
This is a how things are handled at companies that get 500+ applications for each open position and/or are just incompetent. At smaller companies it is very likely that actual humans read a résumé without the keyword matching bullshit.
I’m curious if you have much evidence that this is common? I suspect it’s mostly a myth that is easy to believe (it can feel good to believe it because it means you were removed from the applicant pipeline because of stupid computer reasons rather than because you don’t appear to meet some bar to a human, even if that bar is a stupid thing like prestige) and oft repeated.
I definitely think it’s likely that most people on hn don’t work for companies that do this, but maybe hn just skews toward certain kinds of tech jobs.
It happens in big companies. My team is hiring about 100 people in the next 6 months.
For external candidates, we see less than 30% of all resumes. As with all things, 80% of the rest are junk and another 20% are potentially viable candidates.
It would seem most of the over-confident “it’s a myth” comments in this thread are from folks not familiar with typical processes for hiring in the 100s to 1000s.
You shouldn't be hiring that many. The company might, but the managers and people doing the interviews should be looking at maybe 20 positions per year. They can read résumés themselves and give it some thought.
Who sorts which 20 out of 20,000 resumes go to which managers? Not the managers. Who is “the company”? That’s who is keyword searching/matching/filtering.
The hiring manager. We have got tired of hr rejecting great résumés because they don't understand keywords. They have no clue that qt and qt5 are the same thing, and if we put both keywords on they reject someone with only one. Nor do they get that someone great with c++ is almost automatically great at qt. I'd be happy to teach c++ to someone with rust or Haskell experience if they are willing to learn. (And if it is rust we are open to new work in rust, though we have a lot of legacy c++ you will maintain)
We consider ourselves lucky to get 50 résumés, and if even one isn't obviously disqualified that is a miracle
In college I got the advice that if you couldn’t highlight text on your resume then a bot can’t read it and it might get thrown out. And if that might cause it why risk it?
I think the fact it’s low cost to avoid and high cost if it occurs keeps most people just doing it regardless of whether it’s true or not
Sadly I think this is a bit idealistic. Even in healthcare, doctors can't agree on abbreviations to put into a patient's records, and sadly the state of EMRs (electronic medical records) is a shambles due to this sort of data impedance mismatch.
Come to think of it, the latter is probably why many of us will always have a job.
In my company our HR department actually reads all applications.
This is not true for the largest corporations of course. But that is a solvable problem, just provide a standardized form. You can even adapt it to the needs of the industry in question. This is actually a great boon because you can specifically ask for qualifications. There should be enough budget for that at least.
If you company easily discards applications your HR is either incompetent or it just isn't looking for candidates. But even then you try to build a connection to people applying to you. You may meet them later because they tend to work in your industry.
Again, there are special rules for the largest of corporations although I think HR should be manned enough to have a sensible recruitment process.
So what? You're looking at it one job at a time, and for any one given job it might indeed suck. But collectively, the times when you're unfairly tossed out are balanced by the times when other people are, and you benefit from being in a smaller pool. If you spell it "nodejs," surely some HR person somewhere does too.
It's random, and not based on any job-relevant differences, for sure. But hiring is always going to be pretty random so long as we're pretty bad at predicting who'll be good at any given job. So long as there's no systemic bias, it's fine, and selection by idiosyncratic spelling is at least much less biased than a lot of other irrelevant criteria that get used in hiring.
If a company is choosing which candidates to interview using grep, it's very likely the rest of their hiring process is similarly broken, and quite possibly other aspects of their corporate culture as well. They're actually doing candidates a favour by giving a clear indicator of how shit they are, much like a candidate who turns up to an interview drunk.
If you put an idiot in charge of your recruiting, that is the real problem. A perfect AI that would tell you meant "node.js" when your iPhone's autocorrect changed it to "nodule jesus" won't protect you from the jackass running your hiring. There's no protection from being a moron.
I recall people copy-pasting the job description and hiding it in their resume by making the text invisible just so that automated software would see a 100% keyword match. Bonus here if you use 2-3 alternate spelling for each buzzword (C++ and c++). Is that no longer done?
Applicant Tracking systems, like Bullhorn, actually do a pretty decent job of fuzzy matching and normalising CV/resumés. Of course there is room for improvement and you need to ‘SEO’ your particular CV, but you really should do that for each job application anyway.
The examples are a stretch though. I've never heard of anyone losing an interview due to their spelling of nodejs or computer science. I don't buy that's something that happens often enough to spread FUD over.
I think you might be underestimating the number of applicants that some companies deal with. I used to work for a firm that did analytics and surveying for the recruitment industry. While applicant to hire ratios in the tech industry might be 10 to 1, in many jobs you can easily be looking at 100 to 1. At those scales companies are using automated CV scanning tools and outsourced humans to whittle them down to a set that hiring managers can actually cope with. This is why many companies insist that you enter all the information into a structured form.
My basic rule of thumb is that the more structured and formal the CV submission process is, the worse the applicant to hire ratio will be, and so the more people you are up against.
For those of us in the tech industry, especially later in our careers, that probably isn't an issue, but in many other jobs, and earlier in your carreer, then it's going to be an issue. I remember having to re-enter my CV into ATSs back in 2001 for the big companies hiring graduates because they had so many to get throught, smaller companies just wanted your CVs.
That's already the case. You're much more likely to get hired if you know someone on the inside, if you've been recommended. If you apply through the official channels, you might not even get a response.
But I think an official format for CVs might not work so well because CVs in different fields aren't formatted in the same way. A programmer isn't going to put forward the same kinds of things as a musician or a professor in psychology.
As a hiring manager you’d be inclined to disagree.
If resumes could be standardized and easily searchable for information it would help find candidates much more accurately and quickly without having to read.
Even better would be if there was a verification standard where you could get a blue checkmark on your resume meaning all the information is accurate and not made up bullshit, then you could limit searches only to verified resumes and do less investigation or third party background checking.
That sounds like it pushes hiring managers forward, not industries. You get a benefit out of making it easier to fill open spaces, at the cost of everyone else's unique life experiences.
We have a "verification standard" in the form of diplomas and certificates and neither of those have fixed the hiring issue so far.
If you want to skip background checking, just make people bring their proof of certification with them to the job interview. Of course, those certifications barely mean anything in most fields of work, but it's an exact equivalent of the blue checkmark system you propose.
I get that hiring is hard, but that's why hiring managers exist in the first place. If we used a nicely standardised, automatically validated system, all recruiters and hiring managers would be out of a job. Why pay someone to do that stuff when you could pay a cheap machine learned AI to fill a list of requirements for you?
> If resumes could be standardized and easily searchable for information it would help find candidates much more accurately and quickly without having to read.
I see how that can be useful from the recruiting side of the table, but fail to see how the candidate benefits.
Think back to the first time you wrote a resume. For me and all of my college buddies at least, it was super stressful, because we had no idea what it was supposed to look like and nothing but vague rumors to go on. If there were a straightforward canonical answer for how my resume should be formatted it would have saved me quite a bit of work and quite a bit of stress that I was doing it wrong.
A standard requires people to agree. When I was in college, I was told that a resume was a list of buzzwords and past jobs. 1 page max. That got me quite a few jobs. When I interviewed at Google, people were like "this is it?" when looking at my resume (not that it mattered). I started doing interviews and saw the resumes that ended up at Google and they were all multi-page affairs with details about all the projects you worked on at a job. I started doing that, only to be told "wow, that's a really long resume". (Hint: if your role requires 10 years of experience, applicants are probably going to have a long resume!)
The key point here, I guess, is that a lot of people have told me my resume is in the wrong format, but I still got the job. What's the right format for one employer is the wrong format for another employer. So there is no way to win. I now have no clue how to write a resume, but it also doesn't seem to matter.
(Still bitter about how few people ever look at my Github, though. People put that as a custom field in their automated system, and then the interviewers have nothing to say about my projects!)
You could probably only try to verify local education - for the foreign education not that easy. Even employment (even local) is very hard if possible at all. For example, I work for a large corp X as a contractor. It pays my agency Y. Y pays my umbrella company Z. Z pays me as its employee. There is hardly an easy way to link me to company X yet I've been there for 9 years.
I see what you're driving at -- but that implementing that "verification standard" (i.e. standardized, continually updating 3rd-party background checks) would be a huge undertaking, and a larger than cottage-sized industry in itself.
Way, way more than a matter of parsing a bunch of standardized resume documents.
I wonder if there's any advantage in using fonts and layout that are adversarial to automated resume processing, as in if something fails to scan then perhaps a human is more likely to actually look at it. But perhaps HR just throws those out.
Quaint to think so, but no. If a system is using automated scanning and the document doesn’t scan correctly the recruiter is not going through the ones that didn’t work.
I’d wager in many cases they aren’t even aware of an failure to process a resume occurring.
Also in the eventual human review if you have made yours illegible on purpose, you probably lose any advantage you gained with passing te very low bar of automatic filtering
Does anyone OCR? The systems I've heard of just extract the text from PDFs/docs. Then if some bits cannot be extracted, I was asked to type them myself.
Whether or not it should happen doesn't answer the question of why it hasn't happened. The fact is that the very companies that do automated processing of applications probably would welcome resume standardization, so why didn't they do it yet? It's in their power to propose a standard, settle on it, and require it.
I suspect the answer as to why it hasn't happened is simply because most laypeople applicants would find it too difficult to do something like LaTeX or whatever other thing would be necessary to make this a reality, and also, firms in general suck at adopting new technology.
> laypeople applicants would find it too difficult
> to do something like LaTeX or whatever
Those people where a CV and certificates matter, they already construct a CV in the closed system of LinkedIn or Monster's websites. Nobody needs to use LaTeX, there will be websites where anyone can do it, if anything that's an additional business opportunity.
No, I suspect the actual reason is: Nobody will do it unless the big players (LinkedIn etc.) will adopt it. And the big players don't do it because interoperability, and a CV that can be migrated anywhere, isn't in their interests.
" Reducing someone's life history to a list of educational institutions and employers feels robotic ... but that kind of behaviour shouldn't be encouraged."
Reviewing CVs is boring. They should be made easy for the human who is reading. Imagine being a person in a human resources department going through 100s of these... If there was a good portable format, that process becomes far more efficient and interesting.
A good interview after the resume has been accepted would go into precisely what you are advocating for -- the life history and interpersonal context.
Reality is, however and in my experience, that more and more HR departments fall for promises of AI snake oil. Virtually every employer I apply with has it's own portal, often times asking you to construct your CV anew on their website (often with drop downs that don't include the actual job title or skill). Preferably with a five minute web session and no submit confirmation. Couple this with companies increasingly not responding at all due to fear of being sued, and you'll spend hours on an application and won't even know whether the non-response is due to you, or due to the website trashing the submission. As for CV parsing, this has gone so far as to a big news site over here posting helpful articles on how to beat the system. "Favorite" tidbit: Submit two CV - one for humans, one for dumb algorithms.
Oh and btw, I'm applying at non-tech companies increasingly. Maybe humans isn't perfect either. I mean, I submit proof of 25 years of high grade enterprise IT experience, just for some HR person to ask me where my certificate as an "IT technian" is: they don't know what any of that "weird stuff" means, but to them I lack a generic "IT technician" certificate in early career. A thing that didn't even exist back then! But the latter may just well be a quirk of my own, bureaucratic country.
But overall, this is a hot mess already, and I have often times had the same thought: Why can't we have some sort interchangeable format, much like "geek code", just for all the possible job titles, universities, locations, and companies; with some format-dependent, but free text fields because one can never catch all of them in a fixed list?
> I understand that there are real life problems because companies do use automated processing on applications
I'll always remember but never find the post, where someone just hid tons of keywords in a joke resume and got interviews with an alarmingly high success rate.
I agree they should be read by humans, but (and I feel most comments are missing that point) a standard format would help with that.
Having been reading resumes a lot lately, I always breathe a sigh of relief when people hand in the standard latex template, with maybe only the colors changed. The structure is immediately familiar and easy to navigate. I could use a client app that presents a hypothetical standard data format in such a way. My colleagues in sales or customer success on the other hand seem to prefer other formatting styles; they could look at their candidates in that way.
Note that I'm not arguing that this it the best presentation possible for a CV. I'm just arguing that having a uniform presentation between CVs reduces the mental burden of filtering through them. (And if you want to get into that, might reduce bias, especially if you don't include the image in the presentation)
Unfortunately the lack of a standardised format hasn’t stopped technology from being developed which automatically reviews CVs before they hit a human. The tech is already out there.
The problem we have currently is that some CVs fail that review because they’re in a format that the tools cannot parse. If you’re CV is strong but you’re seeing fewer responses despite having a prettier CV compared to your peers, then that might be the reason. So from your stance standardisation might actually help the industry rather than harm it.
> The problem we have currently is that some CVs fail that review because they’re in a format that the tools cannot parse.
My company (we have software for responding to applicants in real time... think apply... instant conversation with a human recruiter) did a study of about 1,200 resumes and found that 14% of resumes had inaccuracies that would cause them to be screened out... and about 9% of resumes were in a unparsable format. A lot of the screening we see out there is really bad - mostly text search looking for key phrases like specific colleges, specific employers or specific skills. If you are imagining indexing resumes with elastic search and making queries, that may actually be better than state of the art which is usually something that turns into a SQL query.
Given a lot of these tools cannot even parse CVs with inlined tables, the tools I’m imagining aren’t sophisticated in the slightest. But they are depressingly common.
CV or resume analyzed by non-human? Or in an universal format? I hope this will never happen.
How "automatized" way could help there? Unifying the CV/resumes for certain company? Sure, if they are have some form - could be stupid Google Form or more sophisticated like Teamtailor [0] - go ahead. You will have a nice databases or resumes and profiles similar to mini-LinkedIn. While I am fan of automation, I don't see any special incentive here that will profit HR team.
I know hundreds of examples, where CV nor initial job interview didn't exactly has shown how good someone is. Also, I personally had experienced a situation I have completely matched the job offer but after talk during interview, I knew it would be waste of time and mentally it will be a bad choice for next year(s) to join their Acme corp.
Do the JSON databases of applicants will help anyhow to solve the human relation job to find out the best candidate for the position? I am highly skeptical and in my humble opinion it will just "dehumanify" the whole process.
Maybe, it is good if you are looking for some warehouse worker like Amazon? /s
I totally agree but we are fighting the tide (ran a resume, linkedin profile, and interview prep shop for 6 years). Everything in the process is being analyzed, parsed and benchmarked using AI, soon to a more granular level. I even came across some interview analysis software which measured the angle of the person's head along with tone and a bunch of other factors!
> Documents like these shouldn't be automatically processed, they should be reviewed by humans.
Human review is fraught with bias. A lot of attention and column inches are devoted to algorithmic bias these days, but let's not confuse that for evidence that human review is perfect or even better. It's just harder to audit.
I don't think anyone is confusing algorithmic bias for evidence that humans are better. It's the other way around -- the argument has always been that algorithms are biased because humans are biased. Having a systemic bias that masquerades as neutral seems worse than having people in the loop.
When submitting my resume and interviewing for jobs, I'd certainly prefer to be subject to: 1) multiple people with different biases, 2) humans that can change their mind and/or recognize their biases, 3) humans that can explain their rationale, and 4) humans that can grasp and weight eccentric and non-conforming experience. I'm not interested in being ranked in a standardized way, because I'm not interested in being a pure cog in wheel.
I also think people who want standardized resumes and automated processing are vastly underestimating the SEO effect it will have on hiring. If resumes are automated, then many people more aggressive than you will game the system to the Nth degree. The same thing that happens to Google and the internet will happen to jobs: the loudest and spammiest will win, and the good content will go mostly unnoticed and unrewarded.
I think resistance to algorithms is about failure modes. If an algorithm becomes standard and you are an edge case then you are in trouble, much the same way you are in trouble today with companies that have automated customer support and make it hard to contact a human being. Losing your email account is annoying but not being able to have a job is worse. Automated systems are designed to avoid mistakes but are bad at correcting mistakes. I would argue that even with humans involved the hiring process is harder to some people, even discounting bias, just because it is more or less standard but at least they have hope in the variance that remains.
I've seen thousands of resumes... it would be nice to have a baseline of requirements for specific positions. An RFC style guide for human parsing . Too many resumes miss what is important.
> I understand that there are real life problems because companies do use automated processing on applications, but that kind of behaviour shouldn't be encouraged.
Ignoring the web, we already do some formatting with the common Education/Work experience trope. We could arrange resumes arbitrarily just to mess with employers, perhaps? They are already pretty easy to parse.
Knowing life, the end result would be multiple competing standards and weird clans would form along with studies proving that, statistically speaking, standard X people are different from standard Z people.
Edit: Humans are weird. I guess current ecosystem is not as bad as it could be ( even though it does suck ).
Yet, companies use AST tools, which automatically processes the CVs and resumes.
One person that is developing such a tool recommended collage students not to add too much links to their CVs because otherwise, their CV will be flagged as a malicious document.
Downside is when you have to make slight changes each time to better emphasize some aspect of your experience, depending on the job applied for. Recruiting agencies often do this to sex up their candidates chance of winning the job lottery.
A really great resume gets beyond the company names and educational institutions to demonstrate, with concrete numbers, how you improved things at your employer.
Reduced latency by 350 ms
Increased engagement by 18%
Reduced AWS spending by 23%
Scaled from 13 to over 300 virtual machines
That sort of specificity and numerical quantification is what makes a great resume stand out.
Documents like these shouldn't be automatically processed, they should be reviewed by humans. Reducing someone's life history to a list of educational institutions and employers feels robotic even for a software developer's mindset.
I understand that there are real life problems because companies do use automated processing on applications, but that kind of behaviour shouldn't be encouraged.