On one side of the coin, it logically makes sense that ATS systems would help recruiters pare down the list. I'd assume there is some auto-rejection happening, but how sensitive is it (getting rid of clearly unqualified candidates vs. rejecting marginal candidates vs. actually qualified but still rejected due to not being ATS-friendly)? You here about job postings getting thousands of applicants and it seems like it would be impossible to go through those by hand.
On the other side, I see lots of posts from recruiters saying ATS's do not judge resumes and everything is done by the recruiter. These posts seem to be very skeptical of anything saying it will optimize your resume for ATS's because there is no need to actually do that.
Would love more perspective on this issue.
A little background: I have 10+ years of professional experience. Polyglot. Experience with extremely small to medium sized companies.
I re-used my previous resume, which did decently well around 2019 when I was looking for a new position, but with updated info for the newest position I had. Applied to a number positions that I was very qualified for -- crickets. This put me in a bad spot mentally, honestly, as it felt like signals that I'm just not "good enough" for fancy "big tech".
I then updated my resume layout to be more parseable (admittedly this is another variable), then signed up for a service that helps ATS-ify my resume. This involved adding numbers to certain line items, reduce repetition, honestly some good feedback in general. Immediate recruiter messages the next day.
I was extremely averse to using a service like this as it just felt "dirty" or was a sign of my personal ethics "giving in" to some sort of system. Honestly, I still kind of feel that way, but this is the game we play now. I know for a fact I've been auto-rejected from a number of roles that I was 100%+ qualified for because of my resume.
Need to optimize your web site/resume for Google/ATS.
With the amount of applications you receive, the easiest first step is looking for a way to exclude someone. If your ATS can't even parse someone's CV properly, that's a great way to exclude it from consideration.
The next level is keyword matching. If your ATS can't find any of the keywords from your listing in the CV, another easy exclusion.
The whole system is completely bs, for both sides, to the point that personal referrals become more and more important. But if you don't play the game, someone else will, and you lose out.
Why?
Is there some official CV format that applicants are expected to follow, which is guaranteed to be parsable by ATS, so that un-ATS-parsable CV shows incapability of following rules?
If your resume does not parse well (for example, a PDF with two columns can sometimes mean the right column text ends up above the left column text) the recruiter will have outdated / wrong information in front of them when they make the decision to do a screening call or not
I know because my cv had two columns for the longest time, and I kept getting recruiters call me "how's it going at <company I left 6 years ago>" and they had my latest cv
When you have 1,000 applications for a job you don't care much about false negatives (having the system reject candidates you might have wanted). Instead you want the software to give you 100, or even 30 resumes to sift through in the hope of phone screening half of them, interviewing only a few, and getting someone who can do the job as quickly as possible.
I suspect that quite a few ATS were struggling to parse the PDF.
Also, they make the hiring process worse. Know the right keywords to put into a resume and you can be a grand mal idiot, and you'll still get interview time.
I get the other side of it. No HR is staffed to cull through 1000s of resumes, but the systems meant to aid them, are poorly developed and tossed out the door in the name of PROFITS.
Note: Profits aren't a bad thing. PROFITS are, because they ignore everything else for the sake of PROFITS, hence the caps.
A simple example.
How many times would anyone here enjoy uploading a resume, then immediately after, having to either type in the entire fucking thing, or spend 20 mins "correcting" the horrific parsing. If more dev time had been spent the process would be much better for the prospective employee, but they didn't. Companies built half-assed parsing, called it done, and moved on.
Developing these systems into more than just the horse shit out there now would go a long way towards solving both issues.
Source: I've been Senior Technical/Hiring manager for most of my career.
The incredible ease of modern communication allows businesses to show their offer to many more candidates that previously possible, and allows those many more candidates to respond from a greater distance and with more ease.
However, now there is a new problem - too many offers to sort through. It eats up their time with candidates that don't suit their needs.
So, what do they do? They increase requirements and use automated systems to cull the results as much as possible. They get more and more picky. This is an entirely reasonable answer to the problem.
However, now applicants need to work harder to show they're worth considering and avoid the automatic culling.
It becomes an arm race. Employers have incentive to reduce the incoming applications to keep their workload lower, while applicants obviously want to get the job so must make their application get through.
We are in the early phases of a recession. The tech-cession is already in full swing. Inflation is through the roof.
There are very few jobs hiring relative to layoffs. Every optimization, no matter how stupid, matters.
Capitalism rewards the razor thin margins, regardless of how filled with puss and scum those margins are.
> The results returned from through the web app are currently entirely mocked / faked. This means that the results returned are not real and are just for demonstration purposes. This will be implemented with real data results in a future release.
Is this ready for use yet?
Stop shotgunning your resume to dozens of companies. Research a few companies where you might actually want to work. Do this from time to time even if you aren't looking; don't wait until you are desperate for work. And the most important thing is to actually have something to offer. If you do the bare minimum and accomplish very little, you probably aren't going to trick some great company into hiring you with resume tricks.
Yeah, even with individually crafting the resume and cover letter, my response rate was incredibly low during my most recent job search. After doing that for a while, you realize it's not worth the time/emotional investment.
This came up recently somewhere else and my two cents:
1) The best role for LLMs in ATS is in rewriting all resumes consistently in a style which matches the biases / "likings" of the reviewer (Humphry Bogart, Exene Cervenka, Kermit the Frog). Applicants guessing at this merely gives them a better shot at being mistakenly included in the cohort which adopts the preferred voice "naturally". If this "liking" is what resumes are supposed to be judged on then where's the math, it's all a smokescreen for bias innit?
2) Some ATS needs to be the first mover in allowing applicants to download / upload a JSON BLOB of all of their info. I'd recommend cryptographically signing it (and calling it a "service") on download like DKIM, and I think a bunch of things would emerge with that.
A scenario I faced was. - I wrote Java in my resume for the role of Java Developer. But there were other words like J2EE, Gradle, Maven, etc. I possess the knowledge, but I didn't write it. - So, I wanted to detect such keywords and develop something that can highlight such keywords.
Conclusion: - I landed three interviews. 1. Big Service Based Company 2. Startups
I have many friends and acquaintances who tell me what it's like from their perspective. Sure, there are tons of job listings on sites like LinkedIn and others. And yet, for some, the experience seems to be sending hundreds of resumes with little or no feedback and even fewer interviews. Some have resorted to writing short one-page resumes custom-matched to the requirements listed in the job post. In some cases they claim this produces better results. However, we are still talking about what I interpret to be in the 5% response rate.
One of my friends got so frustrated after going through a six-interview hiring process in one case and seven interviews in the other --and not being hired-- that he found himself a job in the oil fields in New Mexico manually recording readings every hour while he lives in a trailer in the middle of nowhere. This, after a long career in various technology fields. He fears age discrimination might have a lot to do with it.
So, yeah, not sure how human-to-human contact happens if people never get a response after sending hundreds of resumes. When we need to hire, my rule is simple: Everyone gets a response. Everyone. Because, on the other side of that resume is a dad, mother, brother, sister, cousin, etc. How would you like to be treated? How would you like others to treat those you love? Well, then, the answer is simple. If you are using AI or mechanized processes, you are treating people like cattle.
Not sure what to think.
It gives values for keywords from resume (got it) Then it finds common words between Job Description and Resume I also see another table of values for the job description
Now what? Do I need to compare the tables? Do the values need to be within certain range?