What Do Applicant Tracking Systems Look For?

What do applicant tracking systems look for? In most deployments, nothing on their own. An ATS stores applications, parses each resume into fields and routes candidates through stages a recruiter defines. Filtering happens only where the employer sets it up: knockout questions, minimum criteria and keyword searches a recruiter runs, plus AI-assisted screening in some products.

What applicant tracking systems look for, step by step

1. Parsing comes first. The system extracts name, contact details, employers, dates, education and skills from the resume into structured fields. This is the stage that affects most applicants: a resume the parser cannot read arrives as a record full of blanks, and a recruiter searching by field never finds it.
2. Knockout questions. Yes or no questions attached to the application form, such as work authorisation, location or a required licence, can reject automatically. This is the most common automated rejection, and it is a rule the employer wrote, not a judgement the software made.
3. Recruiter search and keyword matching. Recruiters search the pool by keyword or Boolean query, and some products show a match indicator against the posting. The terms searched are the terms in the posting, which is why mirroring the posting’s language helps and stuffing unrelated keywords does not.
4. AI ranking, where it exists. Some vendors offer AI-assisted screening or candidate ranking. Adoption varies by product and by employer, and rules such as New York City’s bias-audit requirement for automated hiring tools and the EU AI Act’s treatment of employment systems as high risk shape how it is used. Assume a person still reads the shortlist.
5. Formatting the parser can read. A PDF is fine for nearly every modern parser, and so is a Word document. Multi-column layouts, tables, text inside images, content in headers and footers and unusual fonts are what cause misreads. One column, standard section headings, real text.
6. A human decides. The ATS shows a recruiter a list. For the large majority of roles a person chooses who advances, and the system records the choice. What the software does with that record is in how do applicant tracking systems work.

The part of an ATS the public sees is the posting. Every major system publishes a careers page or job board feed: Greenhouse and Lever expose a public postings endpoint, Ashby a public posting API, Workday a career site per tenant. Those pages carry the posting text the recruiter wrote, which is the same text the keyword search runs against. The category is defined in what is an applicant tracking system, and one system’s public surface is described on the Greenhouse source page.

Where JobsPipe fits

JobsPipe is a jobs data API that collects live postings from LinkedIn, Indeed, Y Combinator, Naukri, Workday, Greenhouse, Workable, SmartRecruiters, Ashby, Lever and Paylocity, returns them as one schema with closure tracking and a ghost score, and includes a free tier of 1,000 jobs a month. It reads the public boards those systems publish, so the postings an ATS exposes are queryable by system:

curl -X POST https://api.jobspipe.dev/v1/jobs/search \
  -H "Authorization: Bearer $JOBSPIPE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "source_or": ["greenhouse", "lever", "ashby"], "job_title_or": ["data analyst"], "posted_at_max_age_days": 14, "limit": 25 }'

Each row names its publisher in sources[0].provider and carries the full description, the text an applicant should be mirroring.