Does ATS Automatically Reject Resumes? What Really Happens

Mostly no. An applicant tracking system (ATS) is hiring software that stores applications, turns resumes into searchable fields and lets recruiters search, filter and sort candidates. In most setups a person decides who moves forward. Automatic rejections do happen, but usually because the employer set a rule: most often a knockout question such as “Are you authorized to work in this country?”, and at some employers a filter on things like credentials or gaps in work history. The popular claim that 75% of resumes are rejected by ATS traces back to a 2012 claim by a company that sold a service for getting past these systems, and no method was given for it.

Below: how these systems handle your resume, what the evidence says, and a checklist for formatting and keywords. (In the UK, Ireland and New Zealand a résumé is usually called a CV. Everything here applies to CVs too.)

The short answer, stage by stage

How applicant tracking systems work

Workday, iCIMS, Greenhouse, Lever and other systems differ, and each employer configures its own, but most applications pass through some version of these five steps.

1. Parsing: your file becomes fields

When you upload a resume, the ATS pulls out your name, contact details, job titles, employers, dates, education and skills and fills in a candidate profile. When that works, recruiters can search and sort you. When it fails, you are not deleted. Greenhouse’s help center says (as of 2026-09) that a resume which fails to parse stays attached to the candidate and the details have to be entered by hand. On a busy role, that may never happen, so your profile sits with empty fields that no search will match.

The same page lists what causes failed parses: files over 2.5 MB, resumes uploaded as images, graphics or word art, tables, headers and footers, contact details inside a header, footer or text box, columned layouts, inconsistent section formatting, shortened job titles such as “Sr. Account Exec”, and company names without an identifier like Inc. or LLC. It is one vendor’s list, but a useful guide for any system.

2. Recruiter search, filters and sorting

Next, recruiters work through the pool. Greenhouse’s guide for candidates says (as of 2026-09) that most recruiters review applications in the order they arrive, and that for roles with many applicants they filter and search, with suggested related keywords. This is where keywords matter. A search does not reject anyone, but if a recruiter searches for “Salesforce” and your resume only says “CRM,” you may simply not appear.

3. Knockout questions

Knockout questions are the clearest case of true automatic rejection. They are yes/no or multiple-choice questions on the application form, and the employer decides which answers end the application. Greenhouse documents auto-reject rules (as of 2026-09) that run on answers to custom application questions. Its example is a job driving a tractor trailer, which needs a Class A commercial driver’s license: applicants who say they do not hold one can be rejected automatically. Typical knockouts cover work authorization or visa sponsorship, a required license or certification, working on site in a given location, and sometimes a minimum number of years of experience.

4. AI match scores and rankings, in some systems

Some ATS products now score or grade candidates against the job. As of 2026-09, Workday describes HiredScore candidate grading as a way to prioritize candidates by comparing job requirements with resumes. iCIMS markets AI that ranks candidates on skills and experience and says recruiters choose how much autonomy to give it. Greenhouse’s candidate guide says (as of 2026-09) that its AI does not score or rank applications. So whether a score exists depends on the employer’s system and settings.

A ranking is not a rejection, but when hundreds of people apply, a low rank can mean nobody reaches you before the role is filled. Some places regulate these tools. New York City requires employers that use automated employment decision tools to have them bias-audited and to notify candidates, and has enforced that rule since July 5, 2023.

5. People make the call

Finally, a recruiter or hiring manager reads the resumes that surfaced and decides who gets a call. Silence usually means you were not chosen, or not reached, in a large pool. It is not proof that a robot threw your resume away.

Where the “75% of resumes rejected by ATS” figure comes from

The number, often phrased as “75% of resumes are never seen by a human,” has a traceable origin. A March 2012 Computerworld article by a CIO.com careers writer opened with the claim that applicant tracking systems end 75% of candidates’ chances of an interview as soon as they apply, and attributed it to Preptel, a job search services company. A CIO article from the month before describes Preptel’s product, ResumeterPro, as a paid subscription that analyzed resumes against job descriptions to help job seekers get through these systems. Neither article describes a study or method behind the 75%.

In 2020, HR professional Christine Assaf followed the citation trail. She found career sites citing a 2014 Forbes piece that itself named no source, and traced the claim back to Preptel, which she reports went out of business in August 2013.

What is known, and what is not

• Known: the figure came from a company selling a service to beat these systems, and the articles that cite it describe no study, sample or method.
• Not known: how many resumes are rejected automatically across all employers. We found no public study that measures it, and the answer depends on how each employer configures its system.
• Also known: automated screening is real at some employers. In the 2021 Harvard Business School and Accenture report Hidden Workers: Untapped Talent, which surveyed 2,275 executives in the US, UK and Germany, 75% of US employers reported using a recruiting management system (RMS), software that works alongside the ATS, and more than 90% of employers used theirs to filter or rank candidates at the first stage. 48% filtered out middle-skills candidates with a gap of more than six months in their work history, and 88% believed that qualified high-skills candidates were screened out because they did not match the exact criteria in the job description.

So the one well-sourced 75% in this story is about how many US employers use screening software, not how many resumes it rejects. And the automatic filtering that does happen comes from criteria employers choose, not from software judging your fonts.

A more recent, much smaller data point: a 2025 survey of 25 US recruiters published by Enhancv found that 23 said their systems do not auto-reject resumes, 2 had set up content-based auto-rejection, all 25 use knockout questions when a posting has them, and 11 said their ATS showed an AI or fit score. Twenty-five people is not a representative sample, but it fits how the vendors above describe their own products.

ATS resume myths, checked

How to make your resume work with any ATS

Formatting checklist

• Use one column and standard section headings: Work Experience, Education, Skills.
• Put your name, email, phone and city in the main body, not in a header, footer or text box.
• Skip tables, text boxes, icons, skill bars and photos for anything that matters.
• Upload a .docx or a text-based PDF (one where you can select the text) and keep it small. Greenhouse will not parse files over 2.5 MB.
• Write job titles in full: “Senior Account Executive,” not “Sr. Account Exec.”
• Use one date format throughout, such as “Mar 2022 – Present.”
• Test it: copy all the text from your file and paste it into a plain text editor. If sections jumble or words vanish, a parser will struggle too.

Keyword checklist

• List the posting’s hard skills, tools, certifications and the job title itself.
• Use the posting’s exact wording where it truthfully describes you. If it says “stakeholder management,” do not only write “worked with partners.”
• Spell out acronyms once: “search engine optimization (SEO).”
• Put keywords inside achievement bullets, not only in a skills list, so the person who reads next sees evidence.
• Do not claim skills you lack. Keywords get you found; the interview checks them.

Use this map for each application:

Posting phrase: [exact words from the job ad]
My evidence: [the role, project or result where I did this]
Resume line: [action verb] + [posting phrase] + [result you can verify]

Example: Posting phrase: “demand forecasting.” My evidence: the quarterly plan I own at my current job. Resume line: “Built a quarterly demand forecasting model in Excel that cut stockouts by [X]% across [N] stores.”

If you use an AI assistant to prepare, this prompt keeps it honest:

Here is a job posting and my resume. List (1) hard requirements that could be knockout questions, (2) the ten most important skills and tools in the posting, and (3) which of those my resume does not mention. Do not invent experience: ask me before suggesting wording for any missing term.

Knockout questions: read first, answer honestly

• Before you start, check the hard requirements: work authorization, sponsorship, licenses, location or on-site days.
• Answer truthfully. A false “yes” may get past the form, but it can surface at interview or in a background check.
• If a question is ambiguous, answer accurately and explain in a free-text box or cover letter if there is one.
• If you clearly miss a stated hard requirement, spend your time on a role you qualify for.

Check your resume against one posting

Because every employer sets up its system differently, no tool can tell you what a particular company’s ATS will do with your resume. What you can check is whether your resume uses the language of the job in front of you. That is what Leepo’s ATS fit score does: paste the job link, Leepo imports the role and company, and the score compares your resume with that one posting, showing which keywords you matched and which are missing. From there, AI resume tailoring suggests changes drawn from your own experience, and you accept only the ones you want.

A fair caveat, given the history above: Preptel sold a version of this idea in 2012. A fit score measures the overlap between your wording and the posting. It is not your odds of an interview, and chasing a perfect score by stuffing in words makes the resume worse for the person who reads it. For a sensible target, see what counts as a good ATS score, and for a step-by-step method, how to tailor your resume.

If you are still not hearing back

• Apply early. Recruiters often work in order of arrival, and 13 of the 25 recruiters in the Enhancv survey said applying early helps.
• Reach a person. A referral, or a short note to the recruiter or hiring manager, gets your name in front of someone directly. In Leepo you can find that person, save them against the application and draft the note, which opens in Gmail for you to send.
• Recheck the fit. If many applications bring no interviews, the roles may be a stretch, or your resume may not show evidence for the skills the postings ask for.

The takeaway: the ATS is rarely the villain. Clear formatting gets your resume read correctly, honest answers get you past the form, and the posting’s own language gets you found. People still make the decision.