How to Get Past the ATS: What US Applicant Tracking Systems Really Do
What an ATS is, and what it is not
Almost every mid-size and large US employer runs applications through an applicant tracking system. The names you will meet most often are Workday, Oracle Taleo, iCIMS, Greenhouse, Lever, SAP SuccessFactors, Ashby, and ADP — plus the built-in systems behind Indeed and LinkedIn Easy Apply.
There is a persistent myth that these systems score your resume and auto-reject it. That is mostly wrong, and believing it leads people to do strange, counterproductive things to their documents.
What an ATS actually does:
- Parses your file, pulling text into structured fields — name, employer, title, dates, skills, education.
- Stores the application in a database recruiters search.
- Filters on the answers to the application form's questions, not usually on your resume prose.
- Ranks or tags candidates by keyword match, as a convenience for the recruiter — who then reads the top of the list.
So the real failure mode is not a robot deciding you are unqualified. It is your resume parsing into garbage, or your form answers knocking you out before a human ever opens it.
The knockout questions matter more than the keywords
Most applications include a short set of screening questions. Answer these wrong and nothing else about your application matters:
- "Are you legally authorized to work in the United States?"
- "Will you now or in the future require sponsorship for employment?"
- "Do you have X years of experience with Y?"
- "Are you willing to work onsite in [city]?"
- Minimum salary expectations.
Answer them honestly and carefully. A mis-click on the sponsorship question will route you to the rejection pile regardless of your background — the correct handling of that pair of questions is covered in work authorization on your resume.
What actually breaks parsing
These are the formatting choices that reliably turn a good resume into fragments:
- Multi-column layouts. A two-column template reads left to right across both columns, interleaving your skills into the middle of your job history.
- Tables. Some parsers read them cell by cell in the wrong order; some skip them.
- Text boxes and shapes. Frequently ignored entirely, along with everything inside them.
- Headers and footers. Your contact details often live here in a designer template — and are often dropped.
- Graphics, icons, and skill-level bars. A five-dot rating conveys nothing to a parser and nothing verifiable to a recruiter.
- Nonstandard section headings. "Where I've Made an Impact" does not map to
experience. Use "Experience," "Education," "Skills," "Certifications." - Dates in unusual formats. Use
Mar 2022 – Presentor03/2022 – Present, consistently. - A scanned image or a design-tool export with the text flattened. Nothing to parse at all.
Single column, standard headings, real text, common font, real bullet characters. That is the whole formula.
Keywords: mirror the posting, do not stuff it
Recruiters search the database for the terms in the requisition. If the posting says accounts payable, NetSuite, and month-end close, those exact phrases should appear in your resume — provided you have actually done that work.
Two rules keep this honest and effective:
1. Use the posting's own words. If they say "stakeholder management" and you wrote "client relations," you will not match. Spell out acronyms once with the expansion — Certified Public Accountant (CPA) — so you match either search.
2. Put them where they belong. In the summary, in the bullet points, in the skills list. Not in a hidden block of white text at the bottom of the page. That trick is detectable, recruiters do look, and it gets applications blacklisted rather than boosted.
Tailoring per posting is the single highest-return habit in a US job search.
File type, file name, and the application form
Send a .docx file unless the posting asks for PDF. Modern parsers handle text-based PDFs well, but older Taleo and SuccessFactors installations still mangle them, and you cannot tell from the outside which version you are facing. Name the file Firstname-Lastname-Resume.docx.
When a system offers to autofill the form from your resume — Workday is the main offender here — check every field it filled. Autofill routinely drops a job, scrambles a date range, or truncates a title. The form data is what the recruiter searches, not your attachment.
And where you have the choice, apply on the company's own careers page rather than through a job board's one-click apply. The application lands in the same ATS with more complete data and, on some boards, with a source tag recruiters treat as lower intent.
What about AI screening?
A growing number of employers layer AI tools on top of the ATS to summarize candidates or rank them against the requisition. This does not change the advice: the tools read the same parsed text. Clean structure, honest keywords, and measurable results are what a language model summarizes well, too. What it does change is the cost of vagueness — a bullet with no outcome in it summarizes to nothing.
Build a single-column, parser-safe resume for free on Monta meu currículo?, then check it against the US resume format guide before you apply.
Frequently Asked Questions
Does the ATS automatically reject my resume?
Usually not. It parses and stores your application and ranks it for the recruiter. Automatic rejections normally come from the screening questions on the application form, not from your resume text.
Should I submit a PDF or a Word file to an ATS?
A .docx file, unless the posting asks for PDF. Word parses reliably across every system; older systems still handle some PDFs badly.
Do keywords in white text fool an applicant tracking system?
No. Hidden text is easy to detect, recruiters check for it, and being caught typically ends the application and can get you flagged at that employer.
Which applicant tracking systems do US employers use most?
Workday, Oracle Taleo, iCIMS, Greenhouse, Lever, SAP SuccessFactors, Ashby and ADP cover most large and mid-size employers, alongside the systems built into Indeed and LinkedIn.