ATS advice has a folklore problem
Applicant tracking systems are real, and some formatting and keyword advice about them is genuinely useful — but the topic has also accumulated a layer of folklore that gets repeated so often it's treated as settled fact, even when it isn't. Some of it is outdated advice from older, less capable systems. Some of it never had much basis to begin with. Believing the wrong version of this advice can cost you real time on precautions that don't matter, or worse, convince you to make your resume worse in service of a myth.
Myth: "The ATS rejects resumes with an AI-style score"
Reality: some ATS platforms do include a matching or ranking feature, but "rejection" in the way this gets described — a hard cutoff score below which a resume is automatically discarded, unseen by any human — is far less common than the myth suggests. Most systems are primarily a searchable database and workflow tool for recruiters, who do the actual filtering themselves using search terms and filters. The system is a filing cabinet with a search bar, not an autonomous gatekeeper making the final call in most cases.
Myth: "You need to match the exact keyword density of the job posting"
Reality: there's no known target percentage, and chasing one leads to unnatural, stuffed-sounding writing that reads poorly to the human who eventually opens the resume — which is a real cost, since a human does eventually read it far more often than the myth implies. The useful version of this advice is much simpler: use the same terms the posting uses for your actual skills and tools (their exact phrasing, not just a close synonym), so a keyword search actually surfaces your resume. That's about accurate matching language, not density.
Myth: "Any use of tables or columns will get your resume rejected outright"
Reality: it depends heavily on the specific system and how the tables or columns are used. Complex nested tables and multi-column layouts genuinely do risk parsing errors on many systems — but a single well-structured table isn't an automatic, universal disqualifier the way this myth implies. The safer and simpler practical takeaway is still to avoid tables and multi-column layouts for anything structural, since the downside risk isn't worth it even where the risk is smaller than the myth suggests.
Myth: "White text keyword stuffing is a clever workaround"
Reality: hiding keywords in white text (or otherwise invisible to a human reader) to game a keyword search is both unreliable — many modern systems extract raw text regardless of color and will surface it plainly to a recruiter, looking exactly as deceptive as it is — and, when it does get noticed, actively damaging to your credibility with the one audience whose opinion actually decides the outcome: the human reader. There's no version of this tactic that has a good outcome.
Myth: "PDF resumes don't work with ATS systems"
Reality: this was true of some genuinely old systems years ago, but it's largely outdated advice today. A properly formatted, text-based PDF (not a scanned image) is generally parsed correctly by modern systems, and PDF has the added advantage of preserving your formatting consistently across different viewers — which a Word document doesn't always do. The exception isn't the file format itself; it's a scanned or image-based PDF with no actual selectable text underneath, which no system can read regardless of file type.
What's actually worth your time
Strip away the folklore and the genuinely useful practices are simple: use standard section headings, avoid tables and columns for layout, keep contact info as plain text rather than in a header, match the posting's specific terminology for your real skills, and save as a proper text-based PDF. That's a short list, and all five items are things you can verify yourself rather than take on faith.
If you want a concrete check rather than general theory, running your resume and a specific job posting through Craftume's ATS Resume Scanner shows you the real language overlap between the two documents — a useful, verifiable signal, in contrast to guessing at a system's internal scoring logic that you can't actually see.
FAQ
Is there any truth to ATS systems auto-rejecting resumes at all? Some systems do support automatic filtering based on hard requirements a recruiter sets (a required certification, a minimum years-of-experience field), but this is a recruiter-configured filter on specific fields, not a hidden, universal scoring algorithm judging your entire resume.
Do different companies really use wildly different systems with different rules? Yes — this is part of why sweeping, one-size-fits-all ATS advice should be treated skeptically. There is no single "the ATS"; there are many different platforms with different parsing behavior, which is exactly why the safest approach is a resume that's simply well-structured by general standards, rather than optimized for one specific system's quirks.
Should I stop trying to include keywords at all, since some of the advice is myth? No — using accurate, matching terminology for your real skills is genuinely useful, separate from the debunked density and stuffing myths. The myth is in the extreme, mechanical version of the advice, not in the underlying idea that matching language matters.
Is a plain black-and-white resume really safer than one with any color? A small amount of color (a heading color, a thin divider line) is not a parsing risk on any system — that myth conflates visual color with structural complexity, which are unrelated. Structure (columns, tables, headers) is what creates parsing risk, not color itself.
How do I know if a specific piece of ATS advice I've read is one of these myths? A reasonable filter: if the advice describes the system as making a human-like judgment call (scoring "fit," penalizing a format for looking unprofessional) rather than doing mechanical text extraction and keyword search, be skeptical of it — that's usually where the folklore creeps in.