There is a genre of advice that says applicant tracking systems are robots that reject 75% of resumes before a human sees them. It is repeated constantly and it is mostly wrong. Understanding what an ATS actually does is the difference between fixing the things that matter and wasting an afternoon on white-text keyword tricks that have not worked in a decade.
What an ATS actually does
An applicant tracking system is a database with a workflow attached. When you upload a resume, it parses the file into structured fields — name, contact details, employers, dates, titles, education, skills — and stores the result alongside the original document. Recruiters then search and filter that database, much as you would filter a spreadsheet.
The important consequence: in the overwhelming majority of configurations, nothing is auto-rejected. What happens instead is quieter and worse. If parsing goes wrong, your record is incomplete, so you do not appear in the searches recruiters run — and a candidate who never appears in a result set is indistinguishable from one who never applied.
The formatting that actually breaks parsing
Parsers are more capable than they were in 2015, but a predictable set of choices still defeats them:
- Text inside images. A resume exported as a PNG or a design file with rasterised type parses as an empty document. If you cannot select your name with a cursor, no system can read it.
- Contact details in the header or footer region of a Word document. Some parsers skip those regions entirely, which loses your email and phone number — the two fields a recruiter needs most.
- Tables and text boxes used for layout. Reading order becomes ambiguous, and a two-column table can interleave your job titles with unrelated dates.
- Invented section headings. "Where I've Made an Impact" is charming and unparseable. Use Experience, Education, Skills, and Certifications.
- Dates in inconsistent formats. Pick one — 03/2021 – 06/2024 or March 2021 – June 2024 — and use it everywhere, including for your current role.
Keywords: matching, not stuffing
Recruiters search the database using terms from the job description. Your resume ranks in those searches when it uses the same vocabulary. That is the entire mechanism, and it explains why both common strategies fail: a resume that never mentions the posting's terminology cannot match, and a resume padded with every term in the industry matches weakly on all of them while reading as noise to the human who eventually opens it.
Skills: Agile, Scrum, Kanban, SAFe, Lean, Six Sigma, PMP, PRINCE2, Waterfall, Jira, Asana, Trello, Monday, Notion, Confluence
Delivered a 14-person Agile programme in Jira, cutting release cycle time from six weeks to nine days.
The second version contains fewer keywords and ranks better for the ones it keeps: the term appears in context, with evidence, where both a parser and a person can use it.
The practical method takes fifteen minutes per application. Read the posting, list the eight to ten terms that appear in the requirements, then check that each one you can honestly claim appears somewhere in your resume — preferably inside an achievement rather than in a list.
What genuinely improves your odds
- Submit a PDF unless the posting explicitly asks for .doc or .docx. Modern parsers handle text-based PDFs correctly and the layout will not shift on the recruiter's machine.
- Use one column for the main content. A sidebar for skills is usually survivable; a two-column layout carrying your work history is not worth the risk.
- Put contact details in the body of the document, on the first line, as plain text.
- Mirror the job title from the posting where it is honest to do so. If they advertise "Data Analyst" and your title is "Insights Analyst", the parenthetical "Insights Analyst (Data Analyst)" is both accurate and searchable.
- Save the file with a human-readable name: firstname-lastname-resume.pdf. Recruiters download hundreds of these.
Things that do not work
- White text on a white background. Parsers extract it, humans see it when the text is selected, and it is treated as deliberate deception.
- Keyword blocks in a hidden footer. Same outcome, same reputation cost.
- Applying repeatedly to the same role with slightly different resumes. Every submission lands in one candidate record and the pattern is visible to the recruiter.
- "ATS-optimised" templates that are visually identical to normal templates. There is no secret format — there is only text a parser can read.
Beating an applicant tracking system is unglamorous. Legible structure, standard headings, honest keyword overlap, and a file that opens cleanly will put you ahead of most of the stack, because most of the stack has not done those four things.