How applicant tracking systems actually read your resume

What an ATS actually does

An applicant tracking system is a piece of hiring software that many companies use to manage the flow of applications. When you submit a resume online, the ATS stores it and produces a structured, plain-text version of its contents that recruiters can search. Some systems then rank or filter the stored resumes based on how well they match the job description.

The step that matters for you as a candidate is the conversion to plain text. Everything the system can search or display to the recruiter comes from that conversion. If the conversion loses a section of your resume, that section is invisible to the recruiter, no matter how well it was formatted in the original document.

The conversion is not a judge of quality. It is a mechanical process that expects resumes to look a certain way: a single reading column, standard section headings, and simple typography. When the resume matches those expectations, the conversion is clean. When it does not, content gets dropped or scrambled.

What reliably breaks a resume

The following features cause the plain-text conversion to fail, or to produce output that is harder for a human to read. Avoid them in your submitted resume:

• Multiple columns. Many systems read left-to-right across a row, so a two-column resume can interleave the two columns into nonsense. A sidebar containing contact details may end up appearing in the middle of a work entry.
• Tables. The conversion flattens a table into a sequence of cells, and the reading order is often not the visual order. A skills table becomes a shuffled list.
• Text boxes and shapes. Many parsers ignore objects that are not part of the main document flow. Contact details in a floating text box are frequently lost entirely.
• Headers and footers. Content in a document header or footer is often excluded from the conversion by design, because it is assumed to be decorative. Never put contact details there.
• Fancy list formatting. Bullets as images, unusual bullet characters, or bullets drawn with icon fonts are usually lost. Use the standard bullet or a simple hyphen.
• Custom section headings. A parser is trained to look for "Experience", "Education", and "Skills". A section called "Where I've Made an Impact" is stored but not categorised, so it does not appear when the recruiter filters by work experience.
• Non-standard date formats. A date range like "Spring '21 – Ongoing" is often stored as raw text rather than a structured date. "March 2021 – Present" or "03/2021 – Present" is safe.
• Very small or very large fonts. Anything below roughly 9pt risks being dropped or misread. Anything above 16pt for body text is usually fine but wastes space.
• PDFs that are really images. Some resume builders export a PDF that contains no selectable text — just a picture of the document. That PDF is unparseable by any ATS. Advoke CV exports real selectable text via the browser's print engine, which avoids this problem.

What works reliably

A resume that parses cleanly usually looks like this:

• Single column. Every section flows top to bottom in the order you want it read.
• Name and contact details in the body of the document, at the top, in normal text (not in a header or a text box).
• Standard section headings: "Professional Summary", "Experience", "Education", "Skills", and any of the other common ones.
• Dates in a common format. "Jan 2020 – Present" or "01/2020 – Present" are both fine. Be consistent within the document.
• Simple, common typography. A single sans-serif or serif family. Bold for emphasis. Italics sparingly, if at all.
• Plain bullets. A filled circle or a hyphen. Not a custom icon font, not an image.
• Reverse chronological order within each section. Most recent first.

Keywords: what to do and what not to do

Modern ATS software matches your resume against the job description by looking for keywords. This is not a secret and it is not a hack — it is how the recruiter finds you when they search their database. You should absolutely include the terms the job description uses, if they accurately describe what you have done.

Where candidates go wrong is not in including keywords, but in how they include them.

• Do use the exact terms from the job posting when they match your experience. If the posting says "PostgreSQL" and you have used PostgreSQL, write "PostgreSQL" not "Postgres" or "psql".
• Do place keywords where they belong. A technology belongs in a bullet point that describes how you used it, or in a skills section, not in a keyword cloud at the bottom of the resume.
• Do make sure every keyword is truthful. If you have read about Kubernetes but never used it, it does not belong on your resume. An interview will expose the gap within two questions.
• Do not hide keywords in white text, tiny fonts, or HTML comments. These tricks have not worked for years, and they are a quick way to be removed from consideration.
• Do not stuff the same keyword into the document a dozen times. If the parser counts keywords at all, it counts occurrences in context. Repetition without context looks like manipulation and reads poorly to any human who reviews the resume.

Use standard section headings

Parsers categorise your content based on the headings they recognise. A resume with a section called "Experience" gets read as work experience. A resume with a section called "Where I Have Worked" usually does not, or gets read as an unknown block.

The safest headings are the ones a recruiter would use themselves:

• Professional Summary or Summary
• Experience, Work Experience, or Professional Experience
• Education
• Skills or Technical Skills
• Projects
• Certifications or Licenses
• Languages
• Awards or Honors
• Volunteer Experience
• Publications

You can use any of these and the parser will categorise the section correctly. The exact wording of the first line of a bullet point, by contrast, is entirely up to you — that is where your writing lives.

Date format, chronology, and order

Within the experience section, list roles in reverse chronological order: the current or most recent first. The same applies to education. This is what recruiters expect and what parsers assume.

Use a consistent date format throughout the document. Advoke CV's editor stores dates in a year-month format internally, which renders as "Jan 2020 – Present" or "01/2020 – Present" depending on the template. Either is fine as long as you do not mix them.

For a gap in your history, you have two options. You can leave the gap as it is, and be prepared to explain it in the interview if asked. Or you can fill it with something truthful — a course, a freelance project, a period of caregiving, a sabbatical you are comfortable describing. What you should not do is stretch the dates on either side to hide the gap. Most recruiters check, and a discrepancy between what you wrote and what a background check reveals is a serious problem.

Practical checklist

1. Check the columns. Look at the resume and ask: if I read strictly top-to-bottom, does every line belong to the section it is in? If not, you have a two-column layout and it needs fixing.
2. Check for text boxes. Contact details and headers should be in the main body. Nothing important should float.
3. Check the section headings. Every heading should be one a parser recognises. Rename the interesting ones.
4. Check the dates. Every date range should use the same format. Reverse chronological order in experience and education.
5. Check for images. Any icon, logo, or graphic that carries meaning is a problem. Icons that repeat adjacent text are usually harmless but add nothing.
6. Check the file format. Submit a PDF with selectable text. If you are not sure whether your PDF has selectable text, open it and try to select a sentence with your mouse. If you cannot select, the PDF is an image and needs to be regenerated.