The 75 Percent ATS Rejection Stat Is Made Up: What ATS Resume Services Can and Cannot Fix

The 75 percent ATS rejection stat has no primary source. Here is what the real research says an ATS resume service can and cannot actually fix.

You have probably seen the ad. "75% of resumes are rejected by an ATS before a human ever sees them" — usually followed by a pitch for a resume rewrite, a keyword scanner, or an "ATS score" tool that promises to fix it. If you're on F-1, OPT, STEM OPT, or H-1B and you've already sent a few hundred applications with little to show for it, that number hits where it hurts: it tells you the problem is invisible, mechanical, and fixable for a fee.

Full disclosure before we go further: F1Jobs sells a job-search service to international students and professionals navigating exactly this situation, so we have a financial interest in you believing paid help is worth it. That is precisely why the 75 percent number needs to be checked before anyone builds an argument on top of it — including us. It doesn't hold up, and the honest replacement number points to a different, more useful diagnosis than "your resume is broken."

Where the 75 percent number actually comes from

Search for a primary source behind "75% of resumes are rejected by ATS" and you will not find one. No government study, no peer-reviewed research, and no named dataset produces that figure. Instead, you'll find the number quoted anywhere from 70 to 88 percent across different articles, always without a citation that traces to actual measurement. The claim is widely believed to originate with Preptel, a resume-optimization vendor that shut down in 2013 — the statistic has outlived the company that likely invented it by more than a decade.

A related claim you'll see alongside it — that 99% of Fortune 500 companies use an ATS — has a more traceable origin, and it's instructive. It appears in a footnote of an actual Harvard Business School and Accenture study, but even there it's attributed to Jobscan, which is itself a resume-optimization vendor. In other words, the industry has been citing itself for years.

None of this means applicant tracking systems don't matter — they clearly do. It means the specific number used to sell you urgency is manufactured, and manufactured statistics tend to point you at the wrong fix.

What the real research says instead

The defensible source here is the Harvard Business School and Accenture report "Hidden Workers: Untapped Talent," published September 2021, based on a survey of 2,275 executives with at least 750 each in the US, UK, and Germany, fielded January to February 2020. Its actual findings are more useful than the 75 percent myth, and somewhat more damning of employers than of candidates:

• 88% of employers agreed their own hiring system filters out qualified high-skills candidates who could do the job — 34% said this happens sometimes, 28% often, and 26% always. For middle-skills roles, the figure rises to 94%.
• 75% of US employers use a Recruitment Management System, and over 90% of those use it to filter or rank candidates rather than simply store applications.
• 48% auto-filter middle-skills candidates who have an employment gap longer than six months — relevant if you took time off between OPT and a new role, or between degree programs.
• 72% of employers reuse or barely modify an existing job posting when creating a new middle-skills requisition, which means the keyword profile of a posting is often stale relative to what the role actually needs.

Read together, this says something sharper than "resumes get rejected." Employers configure their own filtering systems in ways that screen out people who could do the job, and they know it. That's a real, well-documented failure mode — but it isn't the one the 75 percent stat describes, and the fix for it isn't primarily a resume rewrite. It's getting your application in front of the right employer with the right configuration in the first place.

One more figure worth flagging with appropriate caution: a Resume Builder survey of roughly 1,000 business leaders conducted in October 2024 found 65% said they would use AI to reject candidates. That's a stated intention from a company that sells resume products, surveyed at one point in time — treat it as an emerging signal to watch, not an established rate.

The math problem hiding under the resume complaint

If you're an international candidate applying to hundreds of roles with no interviews, the resume is rarely the whole story, because the pool of employers who can even say yes is much smaller than the job board makes it look.

According to NFAP's analysis released November 17, 2025, only 28,277 US employers were approved to hire even one new H-1B worker in FY2025. Against roughly 6 million US employer firms, that's about half of one percent — fewer than 1 in 200 — and 61% of those sponsoring employers hired exactly one person. If your application spreadsheet isn't concentrated among that thin slice, no amount of keyword tuning changes the outcome, because most companies you applied to were never going to sponsor regardless of how your resume parsed.

There is a real counterweight, though, and it matters most if you're on STEM OPT rather than chasing an H-1B this cycle: roughly 1,374,769 US employers were E-Verify enrolled as of December 31, 2025, a pool more than 45 times larger than the H-1B-sponsoring one. E-Verify enrollment is also the baseline requirement for an employer to support your STEM OPT 24-month extension — no sponsorship decision required, just an E-Verify-enrolled employer willing to complete the I-983 training plan. A search built entirely around "does this company sponsor H-1B" ignores a much bigger door that's already open during the STEM OPT window.

Ashby's 2026 Talent Trends Report, released May 7, 2026, adds another layer: applications per hire have roughly tripled since 2021 to over 300, and candidates are roughly 50% less likely to land an interview than five years ago. The funnel has gotten harder for everyone at the same time the sponsoring-employer pool stayed a rounding error of the total market. "500 applications, zero interviews" isn't necessarily a sign something is wrong with you — but it is a sign the applications need to be redirected, not just reformatted.

What an ATS resume service can and cannot fix

If you're weighing whether a checker or a paid rewrite is worth it, this table is the honest answer: yes for the left column, no for the right column. Most "why am I getting zero interviews" situations are weighted toward the right column, which is why a resume fix alone often disappoints.

For the mechanics of getting past parsing specifically, see our ATS formatting guide for international students and the broader case for why a US-style resume needs to do more than pass a scanner.

A better diagnostic sequence than another resume rewrite

Before paying for another optimization pass, work through this in order:

1. Pull your actual numbers. Applications sent, interview requests received, and — critically — how many of those applications went to companies with any history of H-1B sponsorship or, for STEM OPT, E-Verify enrollment.
2. Check the screening questions, not just the resume. Many application portals ask "will you now or in the future require sponsorship" as a hard filter before a human ever opens your file. A rejection seconds after submission usually means this question, not your formatting, ended the process.
3. Run one resume through a parsing check. Confirm your name, dates, and job titles extract cleanly as plain text. This is the legitimate, narrow use case for an ATS checker — a one-time diagnostic, not a recurring subscription.
4. Re-segment your target list. Separate "confirmed sponsors" from "E-Verify enrolled, sponsorship unconfirmed" from "no evidence either way," and weight your next batch of applications toward the first two groups.
5. Only then consider a rewrite. If your resume is clean, targeted, and still generating no interviews from a correctly targeted list, that's the point where professional resume help earns its keep — because the targeting variable has been controlled for.

Skipping straight to step 5 is the most common — and most expensive — mistake candidates make.

Common mistakes

• Buying an ATS score before checking the target list. A 95/100 "ATS score" on a resume you're sending to companies that don't sponsor changes nothing about your outcome.
• Treating every rejection as a resume problem. A same-day auto-rejection is almost always a screening-question filter; a rejection after a phone screen is a different problem, and no resume tweak fixes both.
• Over-optimizing for keywords at the expense of clarity. Stuffing every possible keyword variant into a resume can make it read as noise once a human reviewer opens it — the ATS is not the only audience.
• Assuming a subscription checker needs repeated purchases. Parsing either works or it doesn't; once you've confirmed clean extraction, you don't need to keep re-scanning the same resume.
• Letting a resume writer suggest misrepresenting your work authorization or fabricating US experience to "beat the filter." That crosses from optimization into misrepresentation on a federal application. Don't sanitize your visa status, don't invent employment history, and treat day-one CPT, unverified sponsorship claims, or paying an employer for sponsorship as disqualifying red flags, not shortcuts.
• Ignoring the referral channel entirely. A referral routes around ATS filtering altogether in many companies' processes; see our guide on using referrals to bypass ATS screening as an international candidate for how that works in practice.

Where paid help can actually move the needle

Given all of the above, the honest case for paying for help isn't "we'll get your resume past the ATS." It's narrower: with roughly 1 in 200 US employers sponsoring even one H-1B and applications-per-hire roughly tripled since 2021, the binding constraint for most visa candidates is volume and targeting sustained over months against a moving employer list — a lot of ongoing work to run alone while working, studying, or living under an OPT or STEM OPT unemployment clock.

That's the gap a dedicated team can close: rebuilding your resume, LinkedIn, GitHub, and portfolio so they hold up once a human opens them, then systematically applying and running direct recruiter outreach against a targeted employer list, tracking it all in one place, and preparing you for interviews and negotiation once they happen. None of that guarantees an outcome, and no credible service should tell you it does — it buys disciplined execution of the targeting and volume problem above, on a clock you may not have the bandwidth to run yourself. Before spending on either a resume rewrite or anything else, confirm the diagnosis first against our breakdowns of why 500 applications with no interviews usually means a targeting problem, not a formatting one and how to build a systematic target list of sponsorship-relevant employers.

Frequently asked questions

Is the 75 percent ATS rejection stat real

No. No primary study supporting a 75 percent ATS rejection figure can be located, and the number quoted ranges anywhere from 70 to 88 percent depending on who is repeating it. It traces back to marketing claims from resume-optimization vendors rather than peer-reviewed or government research, most famously a company called Preptel that shut down years ago.

What does the actual research say about ATS rejection instead

The most credible figure comes from Harvard Business School and Accenture's Hidden Workers report, published September 2021 from a survey of 2,275 executives. In it, 88 percent of employers agreed their own hiring system filters out qualified high-skills candidates who could do the job, rising to 94 percent for middle-skills roles.

Is an ATS resume checker worth it for international students

It is worth it for a narrow job — catching parsing errors, missing keywords, and formatting that breaks when a system extracts your text. It is not worth it as a fix for zero interviews after hundreds of applications, because that pattern is usually a targeting and screening-question problem, not a formatting one.

Can a resume optimization service fix visa sponsorship rejections

No resume service can make a company sponsor a work visa if it has never done so, and no formatting change removes a sponsorship-related screening question from an application. What a service can do is help you identify employers who already sponsor and prepare a resume that clears standard parsing so a human recruiter actually sees it.

Why am I not getting interviews even with a well-optimized resume

In most cases the bottleneck is not the resume file itself but who you are applying to. Fewer than roughly 1 in 200 US employers sponsored even one new H-1B worker in FY2025, so if your applications are not concentrated among that group, a technically perfect resume still will not generate interviews.

If you want a second opinion on whether your bottleneck is the resume or the target list before you spend on either, F1Jobs offers a free 30-minute resume and profile audit, with no payment information required to book it.