There is a large amount of folklore about applicant tracking systems, most of it wrong. They do not reject you for using the word 'the' too often. They do not score your personality. They do not need your resume to be a plain text file from 1997.
What they do is mechanical, and it is worth understanding precisely, because it is the part you control.
The four stages, and where resumes die
Stage one, extraction: the file is converted to text. Image-based PDFs yield nothing. Text inside logos and icons yields nothing. Content in header and footer regions is frequently discarded.
Stage two, reading order: the extracted text is put into a sequence. Two-column layouts and layout tables are where this breaks — the parser reads across the page and merges your left-hand skills list into your right-hand job history.
Stage three, section mapping: headings are matched against a dictionary to decide what each block is. 'Work Experience' maps. 'Where I've Made An Impact' does not, and that block may land in an unindexed 'other' bucket.
Stage four, field population and search: mapped sections are parsed into fields — employer, title, start date, end date, degree, skill. Recruiters then search and filter on those fields. A role whose dates failed to parse may not exist as far as the search is concerned.
What we test at each stage
| Stage | Checks |
|---|---|
| Extraction | Text volume and density, image-only detection, header/footer placement, graphic-locked information |
| Reading order | Column count, wide-gap line detection, table and tab-delimited structure detection |
| Section mapping | Recognised heading names for summary, experience, education and skills; duplicate headings; section order |
| Field population | Name, email, phone, location, per-role title and employer, complete start and end dates, timeline gaps |
| Search & ranking | Keyword coverage against a target posting, impact-verb density, skills-section depth |
Checking against a specific job
Without a job description, keyword scoring can only be generic. Paste the posting and the check becomes concrete: the analyzer extracts the required skills, tools, qualifications and ranked keywords from that specific ad, then reports which ones appear in your resume and which do not.
That is the version worth acting on, because keyword coverage is not a global property of a resume — it is a relationship between one resume and one posting.
What we will not claim
We will not tell you your resume is guaranteed to pass any ATS. Different vendors parse differently, employers configure their instances differently, and knockout questions in the application form itself frequently matter more than the document. Any tool promising a universal pass is telling you something it cannot know.
Frequently asked questions
- Which ATS platforms does this reflect?
- The checks target behaviour shared across Workday, Taleo, SuccessFactors, iCIMS, Greenhouse, Lever, BambooHR and Naukri's resume parser — extraction, ordering, section mapping and field population. We do not simulate any single vendor's proprietary ranking.
- Do ATS systems really reject resumes automatically?
- Automatic rejection on the document alone is uncommon. Knockout questions in the form do reject automatically. The far more common failure is invisibility: your resume parsed badly, so you never surface in the recruiter's search.
- How do I know if my PDF is image-based?
- Open it and try to select a line of text with your cursor. If you cannot, neither can the parser. Our checker flags this as a failed extraction check.
- Are fancy fonts a problem?
- Occasionally, through ligature and spacing artefacts on extraction. It is a minor risk compared to columns and tables. Stick to a standard text font for the body and the issue disappears.
- Should I remove all formatting to be safe?
- No. Bold, italics, bullet lists and sensible whitespace all extract cleanly and make the document readable for the human who sees it after the parser. Strip structure, not style.