Do ATS Really Reject 75% of Resumes? What the Evidence Actually Shows

The Claim You Have Seen Everywhere

If you have read anything about resumes in the last few years, you have met this statistic: "75% of resumes are rejected by ATS before a human ever sees them." Sometimes the number is 70%. Sometimes it is 88%. Sometimes it is applied to specific systems, sometimes to all of them. It shows up in LinkedIn posts, resume-service ads, career blogs, and yes — in plenty of articles that are otherwise trying to help you.

It is a scary number, and scary numbers spread. The problem is that this one does not hold up when you go looking for its source. Since the entire premise of ApplyLeaf is that honest resume advice beats fear-based advice, this is a claim worth taking apart carefully — including being fair about the part of it that is true.

Where the "75%" Number Actually Comes From

Here is the uncomfortable part for everyone who has repeated it, us included in the general industry: there is no study behind it.

The figure traces back to marketing material from a company called Preptel, a resume-optimization vendor active around 2012. Preptel's pitch was, essentially, that most resumes get filtered out by applicant tracking software — so you should pay Preptel to optimize yours. The number was a sales argument, not a research finding. Preptel went out of business in 2013. It never published a study, a dataset, or a methodology that anyone can check.

That origin explains a lot. It explains why the figure drifts — 70%, 75%, 88% — depending on who is repeating it and what they are selling. A real statistic from a real study has a fixed value and a citation you can follow back to a sample size and a method. A number that wanders every time it is quoted is a number with no primary source underneath it. When you try to chase the "75%" claim to its origin, you do not arrive at a study. You arrive at a defunct company's old marketing copy, quoted by someone quoting someone quoting someone.

None of this means the people repeating it are lying. Most are passing along a figure they read in good faith. But "everyone says it" is not evidence, and a statistic you cannot trace to a source is one you should not state as fact — which is exactly why you will not find us putting a percentage on it.

What ATS Genuinely Does

To see why the cartoon version is wrong, it helps to know what applicant tracking software actually is. It is far less dramatic than the myth suggests.

An ATS is, at its core, filing and ranking software. When you apply, it does a handful of unglamorous things:

• Parses your resume file, trying to sort your information into fields like name, work history, skills, and education.
• Stores that parsed data in a database the recruiter can search.
• Ranks you against the posting, so the recruiter sees a sorted list instead of a random pile.
• Filters, when a recruiter or the posting has set up specific criteria — a required certification, a location, a knockout question.

Notice what is not on that list: an automatic "reject the qualified candidate" button. The software's main job is to organize applications so humans can get through them, not to secretly delete good people. When a resume gets rejected, the decision almost always comes from a recruiter reading a ranked list, or from a keyword or knockout filter a person configured — not from the software independently judging you unworthy and hitting delete.

This distinction matters because it changes what you do about it. If a mysterious machine were arbitrarily trashing 75% of everyone, the situation would be hopeless. It is not a mystery, and it is not arbitrary. It is a system with understandable, mostly avoidable failure points.

The Real, Avoidable Reasons Resumes Get Filtered

Here is the honest version of the threat. Resumes do get filtered before a human reads them — just not for the reason the myth claims. The common, fixable causes are:

1. Bad parsing. This is the big one. If your resume uses multi-column layouts, tables, text boxes, headers/footers with key info, or text baked into an image, the parser can scramble it. Your "Senior Analyst" title lands in the wrong field, your dates disappear, your skills never get indexed. The information was on the page; the software just could not read it. A clean, single-column, plain-text-friendly layout fixes most of this.

2. Missing the exact keywords a filter is set to. If a recruiter searches their applicant pool for "accounts payable" and your resume says "AP processing," you may not surface — even though you did the exact work. This is not the software rejecting you; it is a search that did not match because the wording differed. Using the posting's own terms for skills you actually have solves it.

3. Failing a knockout question. Many applications include screening questions — work authorization, required certification, willingness to relocate, minimum years in a specific tool. Answer in a way that misses a hard requirement and you can be filtered automatically. That is the system working as intended, and it is worth reading those questions carefully.

Every one of these is about readability and matching, not about a machine having a vendetta. That is genuinely good news, because it means the fixes are concrete and in your control.

Being Fair: ATS Filtering Is Real

It would be its own kind of dishonesty to swing to the opposite extreme and tell you ATS does not matter. It does.

Most employers of any size use applicant tracking software, and ranking and filtering are real parts of hiring. A resume the parser mangles really can sink in the rankings. A posting really can be set to require a specific credential. A recruiter facing hundreds of applications really does rely on search and sort to decide who to read first, which means how well your resume matches the posting really does affect whether a person ever sees it.

So the takeaway is not "ATS is a myth, ignore it." It is narrower and more useful: the specific claim that ATS auto-rejects three-quarters of qualified applicants is unsupported, but the underlying advice to build a clean, matched resume is sound. The fear is invented. The best practices are not.

The Honest Takeaway

You do not need to be afraid of a number that traces back to a defunct vendor's sales pitch. You need a resume that is easy for the software to read and easy for a recruiter to match to the job. Concretely:

• Keep the format parseable. One column, standard headings, real text, no critical information trapped in tables or images.
• Match the posting's language for the skills you genuinely have, so both keyword searches and fast human skims find your fit. The keyword guides by job title show the vocabulary your field's postings typically use.
• Answer knockout questions honestly and carefully, since those are the filters that actually reject automatically.
• Do not stuff keywords or invent experience. It does not fool the ranking, and it collapses the moment a human reads the page or asks a follow-up question.

That is the whole game — no scary statistics required.

If you want help with the matching part, our free keyword finder reads a job description and shows you its must-have versus nice-to-have terms, right in your browser, no signup. And when you are ready to build a clean, matched version of your resume for a specific role, ApplyLeaf's tailoring tool drafts it change by change so you approve every edit — adding only what is true about you. No myths, no fear, just a resume that actually gets read.