The '75% of Resumes Are Rejected by ATS' Stat Is Made Up

If you have spent any time reading about ATS systems, you have encountered this statistic:

"75% of resumes are rejected by ATS before a human ever sees them."

It appears on resume writing service websites, LinkedIn posts from career coaches, ATS tool marketing pages, and job search guides across the internet. It is cited as fact. It is usually the first reason given for why you need to optimise your resume.

It is not real research.

Where the Stat Actually Comes From

The 75% figure traces to a company called Preptel, which sold a resume optimisation service in 2012. The number appeared in their marketing materials as a claimed finding. No study was ever published. No methodology was disclosed. No academic journal reviewed it.

Preptel no longer exists.

The statistic spread because it was striking, it was useful for selling resume services, and nobody checked. It was copied from article to article, from tool to tool, and now appears in places that look authoritative enough that readers assume it was verified somewhere upstream.

It was not. The 75% figure is a 14-year-old sales pitch from a bankrupt company.

The Research That Actually Exists

The most credible study on ATS screening comes from Harvard Business School and Accenture, published in 2021 as part of the "Hidden Workers" report. It surveyed over 8,000 job seekers and 2,250 employers across the United States, United Kingdom, and Germany.

The key finding: 88% of employers said they believe their ATS regularly screens out qualified candidates.

That number is striking — but it is a different claim from the 75% statistic. A few important distinctions:

It is employer self-reporting, not system data. The 88% reflects what hiring managers and HR professionals believe about their systems. It is not a measurement of how many resumes a system automatically rejected. No ATS vendor has ever published rejection rate data by system or configuration.

"Regularly screens out qualified candidates" is not the same as "rejects 75% of applications." A company might receive 500 applications for a role and screen down to 50 without that meaning 375 were unqualified. The filtering may be appropriate for most of those removed and inappropriate for a smaller subset.

It covers all filtering, not just keyword mismatch. ATS screening includes minimum qualification checks (does this person have the required degree or certification?), geographic requirements, and other criteria set by the employer. Not all filtering is about keyword scoring.

The same study asked what caused qualified candidates to be filtered out. The leading factors were:

• 57% of cases: genuine qualification gaps (the candidate did not actually meet the listed requirements)
• 43% of cases: a combination of formatting failures, keyword mismatches, timing (applied after a quota was reached), and volume (not reviewed before the role closed)

What This Means Practically

The honest version of the ATS story is less dramatic than 75% but still real: a meaningful share of qualified candidates are screened out due to preventable issues — formatting problems that break parsing, keyword gaps that cause a low match score, or missing qualifications that could have been addressed in the application.

These are fixable. Fixing them is the point of ATS optimisation.

The problem with the inflated 75% claim is that it creates the wrong mental model. If you believe three quarters of resumes are auto-rejected on keyword grounds, you optimise accordingly — stuffing in keywords, writing for the machine rather than the human reading after the machine. That approach tends to produce resumes that pass screening but fail the next stage.

The more accurate model: most ATS platforms don't reject on a single threshold. They rank candidates relative to each other and surface the top applicants for recruiter review. Your goal is not to clear an arbitrary bar — it is to rank higher than the other applicants for this specific role. That means matching the job description accurately, formatting for reliable parsing, and demonstrating qualifications clearly.

A Note on Vendor Claims

ATS vendors themselves do not publish rejection rate data. Workday, Taleo, Greenhouse, and iCIMS have not disclosed what percentage of applications are filtered at what stage in their systems. The actual rejection rate for any given job posting depends on how the employer configured the system, what minimum requirements they set, and how many candidates applied.

Third-party ATS tools — including PassATS — do not have access to live employer ATS data either. Any tool claiming to know what percentage of resumes "get through" a real ATS is speculating. What these tools measure is keyword gap, format compatibility, and experience alignment — useful signals, but not a simulation of the employer's system. For more on this, see No ATS Resume Tool Simulates a Real ATS — Including Ours.

---

Paste your resume and a job description into PassATS — get your keyword gap, see exactly which terms are missing, and receive a fully rewritten resume that addresses them. The score is a heuristic. The keyword gaps are real.