TL;DR
- An ATS is a search-and-filter tool recruiters use, not a robot that auto-rejects you. In one 2025 survey of recruiters, 92% said their system does not auto-reject resumes on formatting or content.
- Ignore the “75% of resumes are rejected by ATS” claim. It has no traceable source and traces back to a defunct vendor’s 2012 marketing.
- The formatting rules that actually matter are simple: single column, standard headings, real selectable text, standard fonts, no critical info trapped in images, tables, or headers and footers.
- “Always use Word” is outdated. A clean, text-based PDF parses fine in modern systems; the real failure is an image-based file. Follow whatever format the posting asks for.
- Match keywords honestly. Mirror the real language of the job description where it is genuinely true. Stuffing and white-text tricks are visible and get you rejected.
Almost every “ATS-friendly resume” article online is published by a company selling you resume software, and most of them work by scaring you. We are recruiters, so we will tell you the unglamorous truth: an applicant tracking system is mostly a database with a search box. Understanding what it really does, and what it does not, is worth more than any trick. Here is the honest guide.
What an ATS actually does
An applicant tracking system is the software employers use to post jobs, collect applications, and manage candidates through the hiring process. When your resume lands in one, it does four practical things: it parses your document into structured fields (contact details, work history with dates, education, skills), stores you in a searchable database, lets recruiters search and filter that database by keywords and criteria, and helps them track you through the pipeline.
The single most important word there is search. In Jobscan’s 2025 analysis, 76.4% of recruiters said they search and rank candidates by skills. The ATS is the tool they run those searches in. It is not sitting there rejecting people on its own.
That figure comes from a 2025 survey in which recruiters across platforms like Workday, iCIMS, Greenhouse, and Lever were asked directly whether their system auto-rejects. Twenty-three of twenty-five said no. The handful of exceptions used a hard filter for a genuine requirement, like a mandatory license, not for whether you used the right font. As one recruiter put it, the system rejects you if you do not meet the requirements, “not because your resume formatting is off or you didn’t have a secret word in there.”
The myths to stop believing
Two claims dominate this topic and both are wrong or badly overstated.
“75% of resumes are rejected by the ATS before a human sees them.” This number, which also floats around as 70% or 88%, has no traceable primary source. Chase it and it dead-ends at the 2012 marketing of a resume-optimization company called Preptel, which sold the cure to the fear it was manufacturing and no longer exists. There is no study behind it. The drifting percentage is itself the tell. As the survey above shows, the reality is close to the opposite.
“Recruiters spend six seconds on your resume.” This one is at least real, but it is a vendor eye-tracking study from TheLadders, which found six seconds in 2012 and 7.4 seconds when it re-ran the study in 2018, as HR Dive reported. Crucially, it measured only the first keep-or-toss glance while sorting a stack, not the real time spent on a resume a recruiter is actually considering. Treat it as a reminder to make the top third of your resume clear, not as a law of nature.
What is true is that ATS are nearly universal at large employers. Jobscan’s 2025 review found about 98% of Fortune 500 companies use a detectable ATS (the widely repeated “99%” is a rounded version of that). So your resume will almost certainly pass through one. That is a reason to make it machine-readable, not a reason to panic.
The formatting rules that genuinely matter
Here is what actually helps or hurts parsing, drawn from career-center and vendor guidance rather than folklore.
What helps and what hurts ATS parsing
| Helps parsing | Hurts parsing |
|---|---|
| Single-column layout, top to bottom | Multi-column layouts that scramble reading order |
| Standard headings (Experience, Education, Skills) | Creative headings (“Where I’ve Made Impact”) |
| Real, selectable text | Text baked into images, logos, or graphics |
| Standard fonts (Arial, Calibri, Times New Roman) | Decorative or unusual fonts |
| Job title, company, and dates in plain text | Key facts hidden in headers, footers, or text boxes |
| Simple bullet points | Tables and columns for layout |
Source: UC Berkeley Career Engagement resume guidance and applicant-tracking-system parsing documentation, 2026.
UC Berkeley’s career team puts it plainly: avoid headers, footers, text boxes, tables, colors, pictures, and graphics; use standard fonts and standard section titles; and keep your work history in a consistent title, company, dates format. The reason is mechanical. A parser reads a single column of real text cleanly. Put your job title in a two-column table or a text box and it can misread which title goes with which employer, or miss it entirely.
This is the safe-versus-risky split at a glance.
The .docx versus PDF question, honestly
The old advice to “always send a Word document” is outdated. Modern parsers handle a clean PDF as well as a .docx. Berkeley’s guidance is simply to use a Word or Google Doc or a PDF, and to follow whatever format the job posting asks for.
The real failure mode is not PDF, it is an image-based file. A text-based PDF, the kind you get by exporting from Word or Google Docs, stores selectable text the ATS can read. A scanned or image-based PDF is a flat picture of words, and to a parser it can look like a blank page. The test is easy: open your PDF and try to select and copy the text. If you can, an ATS can read it too. If you cannot, rebuild it from a real document. When in doubt, follow the posting’s stated preference, and text-based .docx is the safest default when none is given.
Keyword alignment, done honestly
Keywords matter, because recruiters search by them. But the honest version is very different from the stuffing the “beat the bot” crowd sells.
Align your resume to the real language of the job description wherever it is genuinely true. If the posting says “stakeholder management” and you have done exactly that, use those words, next to the accomplishment that proves it. Context is what gives a keyword weight. A skill sitting beside a real result, in a role where you plausibly used it, reads as credible. The same word floating in a detached “skills” block reads as noise.
What does not work, and actively backfires:
- Keyword stuffing. A wall of terms you cannot back up gets you surfaced for roles you cannot do, and a recruiter drops you in seconds once they read the actual bullets.
- White text and hidden keywords. Hiding keywords in white font or tiny text is the oldest trick and the most visible. ATS extract all text regardless of color, so your hidden block shows up as plain text the moment a recruiter selects all or pastes your resume, and many systems flag background-matched text automatically. Being caught is disqualifying. Do not do it.
The rule: mirror the job’s language where it is true, prove it with a result, and never claim a skill you do not have. That is not gaming the system, it is describing yourself in the words the search will use.
A copy-ready ATS-friendly structure
You do not need a template file. You need a clean structure in real text. This parses reliably and reads well to a human.
[FULL NAME]
[City, State] | [Phone] | [Email] | [LinkedIn URL]
SUMMARY
Two or three plain-text lines: your title/level, your focus, and one or
two proof points. Mirror the target role's language where it is true.
SKILLS
Comma-separated, real skills that match the job description: [Skill],
[Skill], [Skill], [Skill], [Skill], [Skill]
EXPERIENCE
[Job Title] | [Company] | [City, State] | [Start] to [End]
- Accomplishment with a real number and the skill it demonstrates
- Accomplishment showing scope (team size, budget, users, revenue)
- Accomplishment showing an outcome, not a duty
[Job Title] | [Company] | [City, State] | [Start] to [End]
- Accomplishment with a real, verifiable result
- Accomplishment mirroring the target job's key requirement
EDUCATION
[Degree], [Field] | [Institution] | [Year]
CERTIFICATIONS
[Certification], [Issuer], [Year]Recruiter note: everything here is plain text in standard sections, so any parser reads it correctly, and every strong line pairs a real number with the skill behind it. Keep it one column, keep the headings standard, and put nothing that matters in a header or footer. If you want help turning your raw experience into lines like these, our guide to using ChatGPT to write a resume walks through it, and our resume format guide covers the layout choices in depth.
Using ChatGPT to ATS-optimize (without faking anything)
ChatGPT is genuinely useful for aligning your resume to a job, as long as you keep it honest. The move is to have it compare your real resume against the real posting and tell you what is missing, not to invent content. A prompt like this works:
Here is a job description: [PASTE]. Here is my current resume: [PASTE].
Act as a recruiter who fills this kind of role. List the skills and keywords
in the posting that my resume does NOT currently reflect. For each one, tell
me whether my experience plausibly covers it. Only suggest adding a keyword
where my real background supports it, and show me where it would naturally
fit next to an existing accomplishment. Do not invent experience, metrics, or
skills I do not have.The last sentence is the important one. AI will happily fabricate a metric or a skill if you let it, and a recruiter spots invented specifics fast. Used as an honest gap-finder, though, it is a fast way to catch the real keywords you left out. Our full library of ChatGPT resume prompts has more, and the pillar guide to using ChatGPT across your job search puts it in context.
What happens after you hit submit
Here is the actual sequence, so you can stop imagining a robot with a reject button.
The real bottleneck is not the parser. It is volume: popular roles now draw hundreds of applicants, and AI has made that worse, with LinkedIn reporting around 11,000 applications submitted every minute. A machine-readable resume with honest, well-placed keywords is how you get surfaced in that pile and read by a person. For the format fundamentals underneath all of this, our ATS resume format guide goes deeper on the layout mechanics.
FAQ
Do ATS automatically reject resumes?
Mostly no. An applicant tracking system is a search-and-filter database recruiters use, not an auto-rejection robot. In a 2025 survey of recruiters across major platforms, 92% said their system does not auto-reject resumes based on formatting or content. The rare exceptions filter on hard requirements, like a mandatory license. Your bigger risk is not being surfaced in a recruiter’s search, which is why machine-readable formatting and honest keywords matter.
Is it true that 75% of resumes never reach a human?
No. That figure has no traceable source and dead-ends at a resume-software vendor’s 2012 marketing; it drifts between 70%, 75%, and 88% depending on who repeats it, which is a sign it was never real. Nearly all large employers use an ATS, but the system stores and surfaces resumes for recruiters to read, it does not silently bin most of them.
Should I use a PDF or Word document for an ATS?
Either, as long as it is text-based. Modern parsers read a clean PDF exported from Word or Google Docs as well as a .docx. The real failure is an image-based or scanned PDF, which looks like a blank page to a parser. Test yours by selecting and copying the text; if you can, an ATS can. Follow the format the posting asks for, and use a text-based .docx when none is specified.
What makes a resume ATS-friendly?
A single-column layout, standard section headings (Experience, Education, Skills), real selectable text rather than text in images, standard fonts, and no critical information trapped in tables, text boxes, or headers and footers. Beyond formatting, mirror the job description’s real language where it is genuinely true, next to accomplishments that prove it.
Does keyword stuffing help with ATS?
No, it backfires. Stuffing surfaces you for roles you cannot do, and a recruiter drops you the moment they read your actual experience. Hidden white-text keywords are worse: ATS read all text regardless of color, so the block shows up plainly and many systems flag it, and being caught is disqualifying. Match keywords honestly, proven by real results.
The bottom line
An ATS is a database with a search box, not a gatekeeping robot, and the scary statistics used to sell resume software do not survive a look at their sources. Make your resume easy for a parser to read (one column, standard headings, real text, standard fonts), align it honestly to the job’s real language, save it as a text-based file, and you have done everything the system actually rewards. The rest is having accomplishments worth surfacing.
Want a second set of eyes before you submit? Our team will review your resume for free and tell you plainly whether it is clear to both the search and the recruiter behind it.