ATS Resume Guide

What Is an ATS Resume? How an ATS Reads and Sorts It

An ATS resume is a resume that an applicant tracking system (ATS) can read. The software parses your file into fields, keeps it in a candidate record, and lets recruiters search and filter. According to Greenhouse, recruiters still decide who advances. What a given system does depends on the vendor and on how the employer set it up.

To build one, see how to make an ATS resume. This page explains the software itself, using what vendors publish about their own products. Every vendor claim below is limited to the vendor named, and we say when a source could not be read.

What ATS means on a resume

ATS stands for applicant tracking system: the software an employer uses to collect applications, store candidates and move them through hiring stages. "What does ATS mean on a resume?" is a common search, but the term is not something you write on the page. In this guide, "ATS resume" is our shorthand for a resume file that these systems can parse correctly. It is not a vendor term, and no vendor we read defines it.

Systems differ. We read documentation from Greenhouse, Workday, Ashby and iCIMS. Lever's help pages did not load for us, so nothing here describes Lever.

How an ATS works, step by step

The order below is our summary of what the vendors document. Employers can switch features on or off, so not every system runs every step.

Step 1: The application arrives

You upload a file through the employer's form. Greenhouse lists .doc, .docx, .pdf, .rtf and .txt as supported resume formats, up to 100 MB (Greenhouse Support, supported formats, updated August 19, 2026). File limits at other vendors: Unknown.

Step 2: The file is parsed into fields

Parsing means extracting data from a document and converting it into structured fields. iCIMS describes it as technology that "extracts data, usually from a Word or PDF document," and converts it into a structured format, often by filling in a form automatically (iCIMS blog, September 11, 2020). Greenhouse says it "scans an imported resume and auto-fills appropriate fields" (Greenhouse Support, unsuccessful resume parse, updated March 2, 2026).

Parsing is not exact. Workday's administrator guide says "Resume parsing results can vary based on resume format and order of words," and that "Workday enables you to review the resume data" (Workday Administrator Guide, last updated June 23, 2023). It also says fields configured as hidden, Languages and Skills are not auto-filled.

Greenhouse lists what makes a parse fail. One cause is a file over its 2.5 MB limit. Another is obvious placeholder data such as "Company 1" or "First Last". The third is formatting: image-only uploads, graphics, tables, columns, and details placed in headers, footers or text boxes. For layout details, see resume format for ATS.

Step 3: The record is stored

When a parse fails in Greenhouse, the resume "has only been attached to the candidate", and you'll need to "manually input the candidate's details into the fields" (same Greenhouse page). So the candidate record holds the attached file plus whatever fields were filled, by software or by a person.

Step 4: Recruiters search and filter

Greenhouse documents a Full Text Search toggle for searching resumes for keywords, with a snippet under the candidate's name showing how the terms are used in the profile (Greenhouse Support, search resumes for keywords; that page was last updated June 6, 2022, so the screen may have changed). Its Talent Filtering page says "the keyword from your search must exactly match the keyword in the application". A "preferred" term means only one term must appear. A "required" term means all must (Greenhouse Support, Talent Filtering, updated March 2, 2026). The page describes no relevance ranking. For how the wording on your resume meets this, see resume keywords.

Who sets the filters matters. According to Greenhouse's candidate blog, filters are "set and controlled directly by recruiters". The same page says that when a role is popular, recruiters may filter the queue (Greenhouse, What really happens after you apply for a job?, January 8, 2025).

Step 5: Questions and rules can act on your answers

Greenhouse says "Application rules are automatic actions applied to job post questions", triggered when an applicant's answer meets the rule's conditions. One action, Auto-Reject, "uses custom application questions to screen and reject candidates based on their responses" (Greenhouse Support, Application Rules Overview, updated March 2, 2026). Ashby describes global auto-reject rules that "automatically archive candidates based on their answers to your global application questions" (Ashby Knowledge Base). The Ashby page does not say whether anyone reviews a candidate before or after the rule applies.

Both pages tie the rules to answers to application questions, and the employer who configures a rule sets its criteria.

Step 6: Some systems add scores or AI assessments

Where this exists, it is optional and vendor-specific. Greenhouse's Talent Matching uses match scores in one of five categories and groups candidates by match score. The page says "It does not automatically advance or reject candidates" and "Recruiters and hiring managers remain responsible for all hiring decisions" (Greenhouse Support, Talent Matching, updated October 5, 2026). Ashby's AI-assisted application review "assesses candidate resumes against each specific criterion" and returns a status of meets, does not meet, or undecided (Ashby Knowledge Base). Its documentation shows tools for reviewers to flag wrong evaluations and does not mention automatic rejection for this feature.

Step 7: People decide

On its candidate blog, Greenhouse says recruiters review applications and decide "whether to advance it or reject it" (same January 2025 page as above). That is Greenhouse describing Greenhouse. Whether any given employer reads every application is not something vendor documentation can tell you.

At a glance: what each page says

Vendor Parsing Search and filter Automatic rejection Scores or AI Page date
Greenhouse Auto-fills fields; failed parses leave the file attached Full text search; exact-match keyword filters Optional rules on question answers Talent Matching scores, does not reject 2022 to 2026
Workday Fills fields; results vary by format; recruiters can review Unknown (not found in pages we read) Unknown Unknown June 2023
Ashby Unknown (not read) Unknown Optional rules on question answers AI criteria statuses Not shown
iCIMS Defines parsing only Unknown Unknown Unknown September 2020
Lever Unknown (pages did not load) Unknown Unknown Unknown Unknown

"Unknown" means we found no readable official page, not that the feature is absent.

Common myths to avoid, and what the sources say

"An ATS automatically rejects most resumes"

The figure usually attached to this is that 75% of resumes are rejected by an ATS before a person sees them. We searched for a primary source for that number and found none, so its origin is Unknown. It is commonly repeated and unsourced, and we do not use it.

What primary sources do say is narrower. Greenhouse says "Recruiters might configure a job in Greenhouse to auto-reject or auto-advance applications based on very specific rules", and that the filters are recruiter-controlled (Greenhouse blog above). Greenhouse and Ashby both document auto-reject tied to question answers (Step 5). We found no vendor page that reports what share of applications any system rejects.

"A score decides who gets seen"

Some products do score. Greenhouse's Talent Matching does, but Greenhouse says it does not advance or reject anyone (Step 6). Separately, Greenhouse says on its candidate blog that "AI doesn’t score or rank applications"; the same page says some other companies use AI to analyze, score and auto-reject. The two statements sit on pages dated almost two years apart (January 2025 and October 2026) and describe different things, so read each as being about Greenhouse and about the feature it names. For Workday, iCIMS and Lever, whether a score exists and what it does is Unknown here.

"Hidden white-text keywords beat the system"

No vendor documentation we read says this works. Greenhouse's Head of Talent Planning and Acquisition, Ariana Moon, said "the reality is that many companies don't automatically filter out candidates that don't have a keyword in their resume". The post says the tactic "isn't as successful as candidates think" (Greenhouse blog, October 11, 2023). That is one vendor's view, not a test. We found no primary source measuring how often hidden text is detected or penalised, so we make no claim about that.

"PDFs can't be read by an ATS"

Greenhouse lists .pdf among its supported formats (supported formats) and tells candidates "Upload a PDF for best results. If an upload fails, try re-saving your file as a PDF and reducing its file size" (Greenhouse Support, MyGreenhouse FAQ for Candidates, updated November 2, 2026). The same company lists image-only uploads as a cause of failed parsing, so a PDF that is a picture of text is a different case from a text PDF. Other vendors' PDF handling is Unknown here.

Automated screening and the law

This is general information, not legal advice. It is limited to what two official pages say.

New York City's Department of Consumer and Worker Protection says Local Law 144 of 2021 prohibits employers and employment agencies from using an automated employment decision tool unless three conditions hold. The tool has had a bias audit "within one year of the use of the tool". Certain notices have been given to employees or job candidates. Audit information is public (NYC DCWP, Automated Employment Decision Tools). Its FAQ defines an AEDT as a computer-based tool that uses "machine learning, statistical modeling, data analytics, or artificial intelligence", helps make employment decisions, and "substantially assists or replaces discretionary decision-making". The law "applies only to employers and employment agencies that use an AEDT 'in the city'" (NYC DCWP FAQ, dated June 29, 2023).

Whether a specific vendor feature or employer falls under that definition is Unknown to us, and we found no official page that says either way for the products named above.

The U.S. Equal Employment Opportunity Commission has a page titled Select Issues: Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures Under Title VII of the Civil Rights Act of 1964. When we tried to open it on November 12, 2026, it returned a not-found error, so we do not describe its content or say whether the EEOC still stands behind it. If this matters for a decision, check eeoc.gov and ask a lawyer.

Sources

All pages were read on 2026-11-12. Vendor pages describe that vendor only.

Disclosure

atsresumeguide.com is published by the makers of AIHawk, a job application tool. This page does not recommend it or any other product, and the About page gives details.

Last verified: