# ATS Resume: What the Software Really Does, Plus a 10-Point Checklist

- Canonical: https://praxy.me/blog/ats-resume-myth
- Author: Pushpak Teja, Co-founder, Praxy
- Published: 2026-06-10
- Updated: 2026-10-20

> **In short:** 92% of recruiters in one survey say their ATS doesn't auto-reject resumes. Here's what an ATS resume really needs: a 10-point checklist, no keyword tricks.

A good ATS resume is just a clean resume: one the software can read correctly and a recruiter can understand in seconds. An applicant tracking system is a database, not a bouncer. It parses your file into fields, stores you, and lets a recruiter search and sort the pile. It does not silently bin you for writing "led" instead of "spearheaded." Do the hygiene below once, then spend every remaining hour on the human who decides.

**The ATS-friendly resume checklist**

1. **File type:** a .docx or a text-based PDF. If you can't highlight the text in your PDF, the software can't read it either.
2. **Layout:** one column, top to bottom. No tables, text boxes or graphics holding real content.
3. **Contact details in the body**, not in the page header or footer.
4. **Standard section headings:** Experience, Education, Skills, Certifications. Not "My Journey."
5. **Job titles and employers in plain text**, each role with dates in a consistent format (Jan 2023 to Mar 2025).
6. **Skills written as words**, not as icons, star ratings or progress bars.
7. **The job's own terms for work you've actually done.** If the posting says "SQL" and you used SQL, write "SQL."
8. **No hidden text.** No white-font keywords, no secret instructions to an AI screener.
9. **Honest answers to the knockout questions** on the application form. These, not your formatting, are the filters that run automatically.
10. **A top third a human can read in seconds:** your target title, your strongest number, the role's core skills.

Here's what most people get wrong. They believe a robot rejects most resumes before a human sees them, so they pour hours into keyword density and formatting superstition. The real driver is plainer. Your resume either parses cleanly and obviously matches the job, or it gets lost in a pile that is now enormous. Clarity beats tricks on both counts.

## What is an ATS in a resume, and what does it actually do?

An ATS (applicant tracking system) is the software employers use to collect, store and manage applications. "ATS resume" just means a resume that software can read without mangling. Big employers almost all have one: Jobscan, a resume-optimization vendor, detected an ATS at [97.4% of Fortune 500 companies, 487 of 500](https://www.jobscan.co/blog/fortune-500-use-applicant-tracking-systems/), in its 2026 usage report.

It does three boring things. It parses your file into fields like name, contact details, titles and employers. It stores you in a pool a recruiter can search. And it filters or ranks candidates in the ways the employer has configured. That last part is where the fear lives, so it's worth being exact about what gets configured.

The hard filters are usually knockout questions. Greenhouse's own documentation describes auto-reject as a rule where, [based on an applicant's answer to a question, they will automatically be rejected](https://support.greenhouse.io/hc/en-us/articles/360000653472-Auto-reject), and those rules attach to yes/no and select-type questions on the application form. Think "Are you authorized to work in this country?" or "Do you hold this license?" You answered those yourself. Nothing about your font choice is involved.

## Does an ATS automatically reject your resume?

Mostly no, with one clear exception: the knockout questions above. When the resume-builder company Enhancv interviewed 25 U.S. recruiters in late 2025, [23 of them (92%) said their systems don't auto-reject resumes for formatting, content or design](https://enhancv.com/blog/does-ats-reject-resumes/). Only 2 of the 25 had set up content-based auto-rejection. Every one of them used eligibility filters like work authorization or location. It's a small vendor sample, so treat it as a field report, not a census. But it matches how the software is documented.

So where does "75% of resumes are rejected by a robot" come from? Not from research. An essay [republished from The Conversation traces the statistic to a 2012 sales pitch by Preptel](https://phys.org/news/2026-07-wrong-modern-job.html), a resume-optimization company that went out of business the following year and never published a method behind the number. A company selling the cure described the disease. The number stuck anyway.

If you're hearing nothing back, volume is the likelier culprit. Greenhouse's hiring benchmarks, drawn from over 640 million applications at more than 6,000 companies, show the average job drew [116 applications in 2022 and 244 in 2025, a 111% jump](https://www.greenhouse.com/recruiting-benchmarks). Recruiters stop reading long before they reach the bottom of a pile that size. That feels like a robot rejection. It's a person who ran out of hours.

## Where is the ATS myth half-true?

In two places, and both are worth taking seriously.

First, parsing can genuinely fail. Greenhouse's support documentation says a resume can fail to import correctly because it [doesn't have enough information, is a non-standard file type, or includes characters its text parsing technology doesn't recognize](https://support.greenhouse.io/hc/en-us/articles/200721684-Why-didn-t-the-candidate-s-resume-import-correctly). Greenhouse accepts [.doc, .docx, .pdf, .rtf and .txt](https://support.greenhouse.io/hc/en-us/articles/360052218132-Supported-formats-for-resumes-cover-letters-and-other-candidate-uploads) uploads. University career centers give the same layout advice, telling students to [avoid tables, graphics, images or text boxes, which can confuse ATS software](https://career.charlotte.edu/blog/ats-and-guide-resumes-2025). A garbled parse won't usually trigger a rejection, but it can leave your profile with a blank title or a missing email, which hurts you in search.

Second, some employers do filter hard on rigid criteria. Harvard Business School and Accenture's 2021 "Hidden Workers" study surveyed 2,275 executives in the U.S., U.K. and Germany. [88% agreed that qualified high-skills candidates are vetted out because they don't match the exact criteria in the job description](https://www.hbs.edu/ris/Publication%20Files/hiddenworkers09032021_Fuller_white_paper_33a2047f-41dd-47b1-9a8d-bd08cf3bfa94.pdf), rising to 94% for middle-skills roles. The same report found almost half the companies screened out resumes with a work gap of more than six months. That is real automated exclusion. Notice what drives it, though: missing degrees, missing credentials, employment gaps. Not fonts, columns or a shortage of buzzwords.

| Real (fix it once)                                              | Myth (stop worrying)                              |
| --------------------------------------------------------------- | ------------------------------------------------- |
| A scanned or image-only PDF may not parse                       | "PDFs are banned by ATS"                          |
| Contact details in the header can get lost                      | "Any design at all gets you rejected"             |
| Knockout questions auto-reject on your answer                   | "A robot reads your career and scores your worth" |
| Rigid criteria (degree, gaps) can exclude you at some employers | "You need 14 exact keyword matches"               |
| A recruiter searches for terms the job uses                     | "Hidden white-text keywords game the bot"         |

## What is a good ATS score, and how do you check yours?

There is no universal ATS score, so there is no official number for you to check. The employer's ATS doesn't hand candidates a score. Every number you've seen came from a third-party checker, and each checker runs its own model, usually measuring how closely your resume's words overlap with one job description. Jobscan, for example, [recommends a match rate of 75% or higher](https://www.jobscan.co/video-jobscan-tutorial) in its own tool. That is one vendor's target for its own metric. It is not a number any employer's ATS shows you, or necessarily computes the same way.

So a 72 is fine and a 50 is a prompt, not a verdict. A low score usually means one of two things. Either the resume describes your real work in different words from the posting (fix it: use their term for what you did), or you're missing skills the job requires (no rewording fixes that, and stuffing the terms in anyway just moves the problem to the interview). Chasing 100% is how resumes become unreadable. As our piece on why [keyword-stuffing now works against you](https://praxy.me/blog/keyword-stuffing-stopped-working) shows, mirroring the posting word for word makes you look like every other applicant who did the same.

What a checker can usefully tell you is narrower than the number suggests: which of the job's core terms are missing for work you've actually done, which section reads weakest, and what to fix first. Read it as a to-do list, not a grade. To check yours:

1. **Run the parse test first.** Paste your resume into a plain-text editor. If your name, email, titles and dates come out in order, the basics are fine, and no checker can tell you more about that than your own eyes.
2. **If you run a checker, run it once.** Use one that asks for the target job, since a score with no job attached measures nothing in particular.
3. **Fix the findings, not the number.** Act on the top fixes, then stop. A second pass to squeeze out four more points is the keyword-chasing trap above.

## How do you make an ATS-friendly resume that a human also wants to read?

You write for the recruiter, inside a format the software can't break. The format is the checklist above. The writing is where the decision gets made. (For the writing side, run [a resume review that checks target, level and outcomes first](https://praxy.me/blog/resume-review-what-it-should-catch).)

Ladders' 2018 eye-tracking research on recruiters found [an average initial screening time of 7.4 seconds](https://www.prnewswire.com/news-releases/ladders-updates-popular-recruiter-eye-tracking-study-with-new-key-insights-on-how-job-seekers-can-improve-their-resumes-300744217.html), and the resumes that held attention had simple layouts with clearly marked section and title headers. The ones that lost it had cluttered layouts, multiple columns and little white space. The parser and the person want the same thing. That's why we argue [the few seconds your resume gets are real](https://praxy.me/blog/the-six-second-resume-scan-is-real), and why the top third matters most.

Picture two hypothetical product managers applying for the same role.

**Weak (passes any parser, dies with the human):**

> "Results-driven professional with a proven track record of leveraging cross-functional synergies to drive impact."

**Strong (passes any parser, earns the seven seconds):**

> "Product manager, 6 years. Took a checkout flow from 61% to 78% completion by cutting three steps and a forced signup."

Both parse perfectly. Only one makes a recruiter search for nothing else. Keywords land naturally in the strong version because it describes real work in the language of the job. For more rewrites, see [resume bullet points: weak vs strong](https://praxy.me/blog/resume-bullet-points-examples).

## Can ChatGPT make an ATS-friendly resume?

It can format one, and it can also make you invisible. An AI tool will happily produce a single-column resume with standard headings, which is the easy part. The hard part is content only you can supply: the numbers, the named projects, the before-and-after. Left alone, the tools converge on the same polished phrasing, which is why [AI resume tools made everyone average](https://praxy.me/blog/ai-resume-tools-made-everyone-average).

Don't use AI to hide instructions in your resume, either. In Greenhouse's survey of 1,200 U.S. job seekers, [41% admitted using prompt injections, hidden text designed to bypass AI filters](https://www.greenhouse.com/newsroom/an-ai-trust-crisis-70-of-hiring-managers-trust-ai-to-make-faster-and-better-hiring-decisions-only-8-of-job-seekers-call-it-fair), and 22% of hiring managers in the same research said they had caught applicants doing it. Measured rates look far lower: in a 2026 study, researchers who scanned roughly 200,000 real resumes found [approximately 1% contained hidden prompt injections](https://arxiv.org/abs/2605.28999). Either way, a human opens the file eventually. Once they find hidden text, you're no longer a strong candidate with a parsing problem. You're a candidate who tried to cheat.

## Who does this advice fail? (the honest part)

It fails people who don't meet a hard criterion. If the employer's system filters on a degree you don't have, or on a gap longer than six months, no formatting fixes that. The Hidden Workers data says those filters are common enough to matter. Your better route there is a referral or a direct message to the hiring manager, which goes around the filter instead of through it.

It's also thin evidence in places. The 92% figure comes from 25 recruiters interviewed by a company that sells resumes. The Fortune 500 figure comes from a company that sells ATS optimization. Neither is neutral, and every employer configures its ATS differently. Hold the exact percentages loosely. The direction is consistent across all of them: hard filters are explicit and limited, and the rest of the decision is made by people under time pressure.

And the "do it once" advice assumes a normal application. Some high-volume employers do rank by match score, and Enhancv's recruiters said a low score can push you down the pile even when it doesn't reject you. That's one more reason to apply where your experience genuinely fits, rather than spraying a single resume at every listing on [the jobs board](https://praxy.me/jobs).

## What to do now

1. **Test your file once.** Open your PDF and try to highlight the text. Paste it into a plain-text editor and check that your name, email, titles and dates come out in order.
2. **Move everything into one column.** Pull contact details out of the header, and take skills out of icons, tables and text boxes.
3. **Rename your headings to the standard ones.** Experience, Education, Skills. The parser and the recruiter both look for them.
4. **Rewrite your top three bullets as outcomes.** Use the posting's terms for work you actually did, plus one number each.
5. **Answer knockout questions honestly, and skip roles you fail on them.** They're the one filter that runs automatically, and no resume overrides it.
6. **Stop checking ATS scores after one pass.** Once you're in a sensible range, more keywords make the resume worse for the human who decides.

_Not sure whether your resume is being lost to the parser or to the pile? Send it to Praxy on WhatsApp with one job you want. I'll check that it parses cleanly, then rewrite the top third so a recruiter sees the match in the first seven seconds._