How Does an ATS Work?

An applicant tracking system (ATS) does three separate jobs: it parses your resume into structured fields like job titles, dates, and skills; it stores that data in a searchable database; and it lets recruiters search and filter candidates, sometimes with a ranking layer on top. Most "the ATS rejected me" stories are really a recruiter's keyword search that your resume didn't match.

Strip away the mythology and an ATS is a database with a hiring workflow bolted on. Systems like Workday, Greenhouse, Lever, iCIMS, and Taleo exist so a recruiter juggling dozens of open roles can post jobs, collect applications in one place, move candidates through stages, leave notes, and schedule interviews. Screening you is a side effect of that job, not the point of the software.

This matters because the popular picture, an algorithm reading your resume and deciding your fate, gets the mechanics backwards. In most pipelines a human still decides who advances. The ATS shapes which resumes that human sees, and it does that through three distinct mechanisms: parsing, search, and ranking. Advice that makes sense for one is often useless or harmful for another, which is why so much ATS advice online contradicts itself.

One disambiguation before we go further: if you searched this while wiring a generator, you want the automatic transfer switch, a different ATS entirely. Everything below is about hiring software.

The moment you hit submit, the ATS runs a parser over your file. It extracts the raw text, identifies sections (experience, education, skills), and maps what it finds into database fields: name, email, employers, titles, start and end dates, degrees, skills. That structured record, not your carefully designed PDF, is what the system works with from then on.

Parsing is where formatting can genuinely hurt you. If the parser can't read a piece of information, that information does not exist inside the system, no matter how prominent it looks on the page. A recruiter searching the database for "project management" will never find the candidate whose skills live inside an image.

The quick test: open your PDF, select all, and paste into a plain text file. If your bullets come out readable and in order, a parser will handle them. A text-based PDF with a clean single-column layout and standard section headings parses reliably in every modern system, so submit PDF unless a specific employer explicitly asks for something else. If you want a fuller checklist, see how to make your resume ATS-friendly without ugly design, and here's how to check if your resume is ATS friendly in a few minutes.

Worth saying plainly: modern parsers are better than the horror stories suggest. Greenhouse and Lever handle reasonable formatting fine. The failure cases above are avoidable, and avoiding them costs you nothing in design quality.

Once parsed, your record sits in the database alongside every other applicant, for this role and often for past ones too. Here is the step most advice glosses over: for many roles, nobody reads every application top to bottom. A recruiter with hundreds of applicants opens the ATS and searches or filters, the same way you'd search your inbox.

A real query looks something like: "account executive" AND Salesforce AND SaaS. Or the recruiter filters by a required credential, years of experience, or location. Your resume competes at query time. If the recruiter searches "CPA" and your resume only says "certified public accountant," or the reverse, you can lose the match on a technicality.

Two practical consequences follow directly from how search works.

The other search-time mechanism is the knockout question, those application-form questions about work authorization, required licenses, or willingness to relocate. These are the one place where genuine auto-rejection is common: answer "no" to a hard requirement and the system can screen you out before any human looks. That's employer configuration, not the software being hostile, and no resume formatting trick gets around it.

If you're not sure which terms matter for your role, we keep a reference of 800+ ATS resume keywords by job title pulled from real postings.

Some systems add a ranking layer: a match score or sort order based on keyword overlap with the job description, and increasingly on AI-based matching of skills and experience. Recruiters can use it to decide whom to read first. That is its real effect, ordering the queue rather than deleting candidates from it.

Three things people get wrong about ranking:

This is why "beat the ATS" is the wrong frame. You can't outsmart software whose main job is storage and retrieval. What you can do is make sure you parse cleanly, match the searches a recruiter will actually run, and rank high enough to be read early. Those are three different fixes, and now you know which mechanism each one targets.

Everything above collapses into a short, mechanism-by-mechanism checklist.

That last step is the one people skip because it's slow, and it's the one with the most direct payoff. The honest way to speed it up is to see the gaps before you edit. In Roleframe, you duplicate your base resume for a job, paste the posting, and get a fit report: an ATS score against that posting, the exact keywords the job asks for that your resume is missing, and a prioritized plan. Then you make the edits yourself, with Remi, its career copilot, suggesting rewrites you approve line by line. Here's the full walkthrough of checking your resume against a job description and fixing the gaps.