Hacker News Reader

Comment by Aurornis | Hacker News Reader

Blaming the ATS is a popular myth.

There may be edge cases where some obscure software can't parse a weird resume format, but in general if you put anything remotely resembling a common resume format (Company name, dates worked there, optional description and/or bullet points) it will be parsed properly.

I'm in a big Slack where people ask and give career help. Several hiring managers have offered to test people's resumes on their company's ATS over the years. Nobody has ever found a combination that actually failed the ATS except when it was an obvious problem (like someone who made their resume in Illustrator) or an obvious user error (exporting a resume as an image).

View on HN

I wouldn't say it's a "myth". My experience is a lot more like the OP's than what you describe.

A few job sites, like WorkDay, have mangled my PDF resume every time I upload it. Like the OP, I've had to massage the layout and formatting to make it more compatible.

Same. Instantly got two hits back after reformatting. I also simplified my experience section by reducing technicals and focusing more on role responsibilities.

It's not a myth (I wrote ATS systems, and you usually just pay some third party to do the core parsing job as a library or a service and take your failures on the nose) but at the same time most recruiters are NOT going into past resumes and filtering through them, and most of the wide shot resume hits you get on linkedin are for the lowest tier of jobs.

Almost all recruiters are "most recent resume in" type of folks or "curated long term list" type of folks - they'll either circle you for years trying to get you hired for something or they'll forget your name in 25 minutes.

My sense is _most_ resumes are actually reviewed by a human. Most job postings aren't getting 1000s of applicants. If they are, 3 people can review 1000 applications in 2 hours (30s look).

Most companies are not Google. I did some hiring (3 or 4 positions) at a 10 year old stagnant growth “startup” and we got at most 10 applications per day to these engineering positions. This was 2022-2024 so not just the good years.

Same experience from me. Most website will show you the result of the CV parsing and in the end I find it takes me less time to populate the different parts manually on the recruiter's website than to try to correct the mess those systems made with the risk of errors it introduces. I wish there was some kind of json/xml standard for CVs.

I've been applying to jobs for the last three months and many companies use systems that automatically parse your resume and then let you manually correct the parsed data. It's nearly always a mess, and I use a very straightforward linear resume template.

It's possible that there is some selection bias here, where mediocre parsing systems give you the option to manually correct everything because they know they are mediocre. I remain skeptical.

I make my resume in Figma, would that break it?

Would you yourself be able to write a Figma-generated PDF parser that reliably extracts necessary semantic content from rasterized text with random layout? If no, likely an average programmer that wrote the application tracking system wouldn't be able to do that either. If yes, then you are very good, and top companies are hunting for you already, you don't need a Figma resume.

Good point! I actually tried my figma pdf resume with the open parser tool mentioned in the thread and sure enough, complete garbage output.

Reply on news.ycombinator.com