Beating the ATS: How to Get Your Resume Past the Screening Software

You apply. Nothing happens. You apply again, and again nothing happens, and somewhere around application forty you start picturing a machine on the other end that skims your resume for half a second, decides you don't have the magic combination of words, and deletes you before a human ever sees your name. That picture is vivid, it's satisfying in a grim way because it gives the silence a villain, and it's mostly wrong.

The software really does exist. It's called an applicant tracking system, or ATS, and nearly every mid-size and large employer runs one. But what it actually does to your resume is less dramatic and more fixable than the auto-rejection story suggests — which matters, because if you're optimizing for the wrong threat, you can spend hours "beating" a robot that was never going to reject you on its own, while ignoring the formatting choice that actually made you disappear.

The Myth That Won't Die

Ask ten job seekers what an ATS does and at least seven will tell you it scores resumes against a job posting and auto-rejects anything under some invisible cutoff. iCIMS, one of the larger ATS providers in the industry, addresses this directly in its own guidance for employers: an ATS rarely rejects or refuses to process a file on its own, and when it does, it's almost always a technical problem with the upload — a corrupted file, not a mediocre resume. Their explanation is blunt about where the actual decisions get made: "A human is always in control of which resumes are considered and which aren't."

That doesn't mean nothing gets filtered before a person looks at it. Research from Harvard Business School's "Hidden Workers" study, led by Joseph Fuller, found employers automatically filter or rank the large majority of candidates before a human opens the file — 94% for middle-skill roles, 92% for high-skill roles. But filtering and ranking aren't the same act as instant rejection. Recruiters set the rules — required years of experience, a license, work authorization — and everything else gets sorted, not deleted, so they can search and prioritize instead of reading every application cold. Jobscan's 2026 audit of Fortune 500 hiring systems found a detectable ATS at 487 of the 500 companies, or 97.4% — a sign these tools are nearly universal at that scale, not that a formula is silently rejecting you at every one of them.

The distinction matters because it changes what you're actually solving for. You're not trying to trick a scoring algorithm. You're trying to make sure your resume parses cleanly enough, and speaks the recruiter's search language closely enough, that when they go looking for someone like you, you show up.

What the Software Is Actually Doing

Picture a library card catalog with a bad habit: it tears pages out of any book bound in a way it doesn't recognize. That's closer to reality than "bouncer with a checklist." When your resume lands in an ATS, the system strips out the text, sorts it into fields — name, employer, dates, skills — and indexes it so a recruiter can search it later, the same way you'd search a database. A recruiter doesn't read four hundred resumes top to bottom. They run something closer to a query: find people with this title, this certification, this range of experience.

If the extraction step goes wrong, everything downstream goes wrong with it, silently. Your resume isn't marked "rejected." It's marked "unreadable," or populated with garbage in the wrong fields — and it never surfaces when a recruiter searches for exactly what you have.

Where Parsing Actually Breaks

Formatting doesn't fail an ATS because the software is unsophisticated about design. It fails because the parser is trying to answer one question — "what is this text and where does it belong" — and certain layout choices make that question unanswerable.

None of these choices look risky on your screen. They look risky to a parser that has no idea what a text box is, only that the characters it expected in a particular position aren't there.

Matching Keywords Without Gaming Anything

The search half of an ATS matters as much as the parsing half, and it works less like a slot machine and more like a search bar. Recruiters build queries out of the job description's actual language, and the system surfaces resumes containing that language — sometimes exact matches, increasingly close variants and synonyms as the tools improve. Stuffing a resume with every term from the posting doesn't outsmart that; it just produces a resume that reads badly to the human who eventually opens it, and that human still makes the call.

Don't write: "Results-driven professional with expertise in SEO, SEO strategy, search engine optimization, and SEO campaign management seeking SEO role."

Write instead: "Led SEO strategy for a 40-page product catalog, increasing organic traffic 34% in six months through keyword research, technical audits, and content restructuring."

The second version contains the term the job posting cares about, uses both the acronym and the spelled-out phrase somewhere in the document, and still reads like something a person wrote about work they actually did. That's the real target — not maximum keyword density, but honest overlap with the employer's vocabulary, stated the way a human would say it.

PDF or Word? The Actual Answer

This question gets more contradictory advice than almost anything else in resume writing, and the honest answer is: it depends on where you're sending it, and it's less binary than either camp claims. Jobscan's own testing recommends PDF as the default — "convert your Word resume into PDF when applying for a job, unless the job posting specifically requests a .docx file" — because most current systems parse PDFs accurately and PDFs preserve your formatting exactly as you built it, so what a recruiter sees on screen matches what you designed. That's the right default for the large majority of applications you'll submit in 2026.

The nuance is in the exceptions. Some older, rigid systems — still common in government hiring, unionized industries, and legacy enterprise setups — parse Word documents more reliably than PDFs, which is why some postings tell you explicitly which format to upload. Follow that instruction over any general rule; it exists because that employer's system has a known preference. Either way, confirm your file actually contains real text: a resume built by scanning a printed page or exporting a flattened image has no text layer, and no parser can read it. Try to select and copy a sentence from your file. If you can't, neither can the system.

Headers That Parse the Way You Intend

Section headers do double duty — they're what a human scans for first, and they're the anchor words a parser uses to sort your content into the right fields. Creative headers might read as more personable to you, but to a parser they're just unrecognized text, which means everything underneath them can end up unsorted too.

Keep each header on its own line, in the main body of the document, in plain text — not inside a table cell, not styled as an image, not tucked into a header or footer that might get stripped before it's even read. Boring section names are doing real work here. The parser doesn't reward creativity; it rewards being recognized.

Run through that list before every application, not after you've already sent forty resumes with a file the parser couldn't read.

The Point of All This

None of this is about defeating an adversary. There's no formula to reverse-engineer and no algorithm keeping score against you — there's a recruiter, searching a database, using software with real limitations, trying to find people who can do a specific job. Your resume's only task is to be readable by that software the same way it's readable by that person: plain formatting, honest language, standard structure. Get that right and the silence stops being mysterious. You're not up against a machine. You're up against a search bar, and search bars just need to find the right words in the right place.

This is really the mechanical half of the problem — for a closer look at the formatting choices that most often cause a resume to get misread before a recruiter ever sees it, and once the resume itself is clearing that bar, a cover letter that actually gets read is usually the next thing worth fixing.

About the author

Sophie is a byline for Jobssfusion's research and editorial team with 3+ years of experience covering job search strategy, salary and compensation. Content is built from primary sources BLS and DOL data, ATS provider documentation, and published labor-market research — with every claim linked so you can verify it yourself.