Chapter 04 · Verifiable

Get Past the ATS Without Sabotaging the Human Read

You have probably heard that 75% of resumes are rejected by the applicant tracking system before a human ever sees them. That number is fake. When a 2025 study asked 25 recruiters across tech, healthcare, finance, and six other industries whether their ATS auto-rejects resumes on formatting or content, 92% said no1.

What almost every recruiter does use the system for is filtering2. Filtering is not rejection. But the layout that survives the filter is also the one a human can skim, and most job-seekers accidentally break both.

By the end of this chapter, you will have a resume in a one-column, plain-text-friendly format that a Greenhouse-style parser can read cleanly3 and a human recruiter can scan in about seven seconds without losing the plot4.

An adorable robot skims a paper with a suspicious look
Fig 4.1 — What a resume parser looks like reading a two-column layout with icons. via GIPHY

Two Readers, One Layout

lightbulb

lightbulb The big idea

Your resume is read twice, in order. First a parser turns your PDF or DOCX into structured fields (Name, Email, Skills, Company, Title, Dates) and files them into an ATS record3. Then a human recruiter opens that record and skims the original document itself.

The parser doesn't judge your resume. It just extracts. When it can't extract cleanly, because of a two-column layout, a table, an icon-based skills bar, or your name sitting inside a page-header block, the record it hands the recruiter is missing the very fields the recruiter searches on3.

And 97.8% of Fortune 500 companies run their careers pages on one of these systems5, so opting out is not an option.

The human read is short. The most-cited industry estimate, from a 2018 eye-tracking study, is 7.4 seconds of initial attention per resume4. Nothing in that window rewards typographic cleverness.

Both readers want the same things: your target job title, your relevant skills, your recent employers, and a couple of numbers. Both readers punish the same things: multi-column layouts, embedded graphics, and any format that hides the words behind decoration.

The good news is that a layout the parser reads cleanly is also the layout a recruiter can skim in seven seconds. You are not optimizing for two audiences. You are optimizing for one shape that serves both.

The 75% Myth, and What Actually Happens

The "75% rejected by ATS" claim has no traceable methodology behind it. What is verifiable, from actual recruiter interviews conducted in late 2025, is the opposite: 23 of 25 recruiters said their ATS was not configured to auto-reject on formatting or content at all1.

The two exceptions had strict, disclosed rules ("match less than 75% of required skills," "fewer than 7 of 10 must-have skills present"). That is not the invisible black-box rejection the myth implies.

What is really happening is closer to this. Almost every recruiter, 99.7% of the 384 surveyed in Jobscan's 2025 report, uses the ATS to filter a large candidate pool down to a shortlist2. Three-quarters of them filter by skills first; more than half use job titles as keyword filters2.

The resumes that don't clear those filters aren't rejected. They're just not surfaced.

The distinction matters because it changes what you optimize for. You are not writing a resume to trick a rejection algorithm that mostly doesn't exist. You are writing a resume that shows up in the recruiter's filtered results and, once it does, survives seven seconds of human attention.

Harvard Business School's Hidden Workers study makes the size of this filtering effect concrete: 88% of employers agree qualified high-skills candidates are vetted out because they don't match the exact criteria in the job description; the number climbs to 94% for middle-skills roles6.

The filter is real. Beating it is not about beating a robot. It is about naming the same things the job description names, in a format both readers can extract.

What the Parser Sees vs. What You Sent

Take Ravi Patel, a mid-career mechanical engineer applying to a manufacturing lead role. He built his resume in a template he liked: two columns, a sidebar with icons for each skill, his name and contact info sitting inside the document header, and a small photo of himself top-right. It looked designed. He was proud of it.

He got no callbacks in six weeks.

The parser at the other end of his submission was doing one job: extract a name, contact info, and a stack of company / title / dates blocks into structured fields.

Greenhouse's own documentation lists five things that reliably break that extraction: graphics and photos, contact info trapped in a header or footer, tables, columned layouts, and any resume uploaded as an image rather than a document3. Ravi's resume had four of the five.

The record the ATS handed the recruiter had a blank Name field: Ravi's name lived in the page header, which the parser skipped. It had a Company field that read "Data Science," which was actually his sidebar's skills header, misread as an employer once the columns collapsed.

And it had a Skills list with three of his nine actual skills. A recruiter searching "mechanical engineer, GD&T, Solidworks" never saw him. His profile didn't match, not because he wasn't qualified, but because the fields the search ran against were empty or wrong.

I have made the same mistake. Two years into my career I paid a designer $200 for a two-column resume with a colored sidebar and small icon glyphs next to each skill. I applied to fifteen roles and got one reply.

When I finally pasted the PDF into a plain-text converter to see what the parser was seeing, my work history had merged into a single unreadable paragraph with the sidebar text interleaved line-by-line. Every application I had sent for two months had landed in the ATS as gibberish.

Ravi's fix took an hour. Single column. Contact info in the document body, not the header. No sidebar, no icons, no photo. Section headings in ALL CAPS so the parser could find the boundaries7. Skills listed as plain text separated by commas.

Nothing else changed. Same content, same bullets, same numbers. Callback rate went from zero in six weeks to three in ten days.

ATS field Ravi's original resume Ravi's rebuilt resume
Name Blank — lived in the page header, which the parser skipped Extracted correctly — moved into the document body
Company Read "Data Science" — his sidebar's skills header, misread as an employer Extracted correctly — no sidebar to misread
Skills 3 of his 9 actual skills All of them, listed as plain text separated by commas
Callbacks 0 in six weeks 3 in ten days

Five Steps to a Two-Reader Resume

1. Move to a one-column layout. No sidebars, no tables, no text boxes. Contact info in the document body directly under your name, not in the page header. This one change fixes the majority of parse failures37.

If you built your resume in a template, the fastest fix is to open a blank document and rebuild the content in a single flowing column. Your bullets are already written. You are only changing the container.

2. Strip the decoration. No embedded graphics, no photo, no icon skill-meters, no colored shading behind section headings. Use one standard font (Arial, Calibri, or Times New Roman) at 10–12 point8.

ALL-CAPS section headings ("EXPERIENCE," "SKILLS," "EDUCATION") give the parser unambiguous section boundaries7. Spell out acronyms on first use so a filter search on the spelled-out form still matches. This is not a design chapter. It is a legibility chapter.

3. Mirror the job posting into your Summary and Skills. Open the job description. Copy the exact phrases in the "must have" section, for example "SQL," "Python," "root-cause analysis," "cross-functional program management," and make sure each one appears verbatim in either your Summary line or your Skills list2.

Do not paste the whole job description; recruiters see the same trick every day. Weave the keywords into the bullets they belong to; a skill listed in an achievement outranks one buried in a comma-separated list7.

4. Save as PDF and DOCX, then send whichever the posting asks for. Greenhouse, the only major ATS with a public candidate-facing help doc on this, accepts both without stating a preference9.

Neither is universally "safer." The single defensible rule: if the posting names a file type, send that one exactly. If it does not, PDF preserves layout across recruiter monitors more reliably than DOCX. Save both versions so you can send either without reformatting.

5. Run the file through a free checker before you apply. Two options that will each score your resume against the specific job posting: Jobscan, which returns a percentage match rate and recommends 80% or higher10; and Resume Worded, which scores across 30+ ATS-compliance and impact checks and treats 85+ as strong11.

Do not rely on Microsoft's Resume Assistant. Microsoft discontinued it in August 202212. Any tool that opens your resume in your browser and reads it back to you in plain text tells you what the parser is going to see.

The ATS-Safe Skeleton

Rebuild your resume in this exact order. No sidebars, no tables, no icons.

YOUR NAME (14–16pt, bold, top of page)
City, State · phone · email · linkedin.com/in/handle
   ↑ in the document body, not the page header

SUMMARY  (ALL CAPS section heading, 12pt bold)
Two lines. Target role + 2–3 signature skills + one metric.
Mirror keywords from the job posting verbatim.

EXPERIENCE
Company · City, State · Month YYYY – Month YYYY
Job Title
· Achievement bullet with a number
· Achievement bullet with a number
· Achievement bullet with a number

SKILLS
Comma-separated list. Include the spelled-out form of every acronym.

EDUCATION
Institution · Degree · Year

Save as LastName_FirstName_Resume.pdf and LastName_FirstName_Resume.docx. Send whichever the posting asks for; if silent, send the PDF.

Seven Checks Before You Send

  • One column, from top to bottom.
    No sidebars, no tables, no text boxes3.
  • Contact info in the document body
    , not in the page header or footer3.
  • No embedded images, icons, or shaded blocks.
    Standard font, 10–12pt8.
  • Section headings in ALL CAPS
    so the parser can find them7.
  • Every "must-have" phrase from the job posting
    appears verbatim in your Summary, Skills, or a bullet2.
  • Both .pdf and .docx saved.
    Send the file the posting names; default to PDF otherwise9.
  • Ran through Jobscan or Resume Worded
    against this specific posting; match rate ≥ 80%1011.

The parser lets you through. The recruiter now has seven seconds. What they read in those seven seconds is a stack of bullets. Every bullet is either doing the work of getting you interviewed, or it is a sentence they skim past. That's Chapter 5.

Notes

  1. Enhancv, Doroteya Vasileva CPRW, "Does the ATS Reject Your Resume? 25 Recruiters Explain What Really Happens," 2025. Interviews with 25 U.S.-based recruiters/TA professionals across nine industries, Sept–Oct 2025. Small sample; treat as directionally strong. https://enhancv.com/blog/does-ats-reject-resumes/

  2. Jobscan, Sydney Myers, "The State of the Job Search 2025." Survey of 384 recruiters, Feb–Mar 2025. https://www.jobscan.co/state-of-the-job-search

  3. Greenhouse Support Center, "Unsuccessful resume parse," accessed 2026-08-14. Vendor documentation naming graphics, headers/footers, tables, and columned layouts as parse-failure causes. https://support.greenhouse.io/hc/en-us/articles/200989175-Unsuccessful-resume-parse

  4. TheLadders via PR Newswire, "Ladders Updates Popular Recruiter Eye-Tracking Study," 2018. The 2018 update to the widely quoted 2012 six-second figure; sample size for the eye-tracking portion is not disclosed in the release, so cite as "the most-cited industry estimate," not a rigorously verified fact. 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

  5. Jobscan, Kelsey Purcell CPRW, "2025 Applicant Tracking System (ATS) Usage Report." Manual detection across all 500 Fortune 500 companies. https://www.jobscan.co/blog/fortune-500-use-applicant-tracking-systems/

  6. Harvard Business School / Accenture, Joseph B. Fuller, Manjari Raman, Eva Sage-Gavin, Kristen Hines, "Hidden Workers: Untapped Talent," Sept 2021. Survey of 8,000+ hidden workers and 2,250+ executives across the U.S., U.K., and Germany. https://www.hbs.edu/managing-the-future-of-work/research/hidden-workers-untapped-talent

  7. Ohio Northern University, Polar Careers, "A Guide to Adapting Your Resume for the Applicant Tracking System (ATS)," accessed 2026-08-14. Institutional career-services guide. https://my.onu.edu/sites/default/files/applicant_tracking_system_resume_guide.pdf

  8. Santa Clara University Career Center, "Common ATS Resume Formatting Mistakes," accessed 2026-08-14. Reprint of Jobscan guidance; corroborates one-column, standard-font, no-headers/footers guidance from a second institutional source. https://www.scu.edu/careercenter/toolkit/job-scan-common-ats-resume-formatting-mistakes/

  9. Greenhouse Support Center, "Supported formats for resumes, cover letters and other candidate uploads," accessed 2026-08-14. Vendor documentation confirming .doc, .docx, .pdf, .rtf, and .txt are all accepted without a stated preference. https://support.greenhouse.io/hc/en-us/articles/360052218132-Supported-formats-for-resumes-cover-letters-and-other-candidate-uploads

  10. Jobscan, "Free Resume Scanner" and "What Jobscan Match Rate Should I Aim For?" Vendor tool with a paid upsell. https://www.jobscan.co/resume-scanner

  11. Resume Worded, "Resume Scanner - Get a Free ATS Resume Scan." Vendor tool. https://resumeworded.com/resume-scanner

  12. Microsoft Support, "Resume Assistant is no longer available." Discontinued August 9, 2022. https://support.microsoft.com/en-us/office/resume-assistant-is-no-longer-available-444ff6f0-ef74-4a9c-9091-ffd7a9d1917a