What an ATS actually does with your resume

There is a large industry built on the idea that an applicant tracking system scores your resume and silently bins it. That is mostly not what happens, and the advice built on it — keyword stuffing, white text, rigid templates — ranges from useless to actively harmful.

What it actually is

An ATS is a database with a hiring workflow attached. Greenhouse, Lever, Ashby, Workday and the rest exist so a recruiter can see every applicant for a role, move them through stages, leave notes and not lose anyone. The parsing step reads your resume and pre-fills structured fields: name, contact details, employers, dates, education, skills.

In most configurations, nothing automatically rejects you. A human opens the list and reads. What the parser affects is how you look in that list, and whether you turn up when the recruiter searches for a skill.

What genuinely breaks parsing

• Multi-column layouts. The parser reads in document order, and a two-column design can interleave your job titles with your skills list into nonsense.
• Text inside images or graphics. Invisible to a parser. A resume that is one exported image parses to nothing at all.
• Text boxes and headers/footers. Frequently skipped entirely — contact details in a header can vanish.
• Unusual date formats. 'Mar '24 – present' is often read worse than '03/2024 – Present'.
• Creative section titles. 'Where I've Made Impact' is not 'Experience' to a parser looking for the latter.

What doesn't matter as much as you've been told

A tasteful single-column PDF is fine nearly everywhere — the long-standing advice to always send .docx is out of date. One page versus two is a readability question, not a parsing one. And keyword density is not a real mechanism: search either finds the term in your resume or it doesn't, so use the words that are true of you and skip the ones that aren't.

White keyword text behind a white background is worth naming separately. It is visible to anyone who selects the text, and it is read as an attempt to deceive. People have lost offers over it.

The one check worth doing

Copy your resume PDF's text and paste it into a plain text editor. What you see is roughly what the parser sees. If your job titles are jumbled, your dates are scattered or your contact details are missing, fix the layout — that is the entire test, and it costs a minute.

Then spend the time you would have spent on template shopping writing sharper bullets. A parser has never decided to interview anyone.