How to pass ATS resume screening in 2026 (what actually matters)

Type "how to pass ATS" into any search engine and you'll drown in advice built on a myth — that a robot reads your resume, scores it against a keyword list, and rejects 75% of candidates automatically. That 75% figure has no study behind it: it traces to a 2012 sales pitch by Preptel, a resume-optimisation company that shut down in 2013 without ever publishing a methodology. It has been repeated ever since by sites selling resume services.

The best available counter-evidence is small but pointed. In a study run in September–October 2025, Enhancv interviewed 25 US recruiters at companies ranging from 100 to 50,000+ employees: 92% said their systems do not automatically reject resumes on formatting, design, missing keywords or a low AI match score. The 8% who did use auto-rejection applied it only to roles with highly specific hard requirements. Twenty-five recruiters is a small sample and worth treating as indicative rather than definitive — but it is considerably more evidence than the 75% claim has ever had. What actually filters you out is a combination of knockout questions and a human recruiter scanning your resume for six to eight seconds.

Start with the knockouts, because they end more candidacies than keywords ever will. When an application asks "Are you authorized to work in this country?", "How many years of experience do you have with X?", or "Are you willing to work on-site?", your answers are filterable fields. Answer them honestly — but before you apply at all, check whether you clear them. If a role wants 8 years and you have 4, your time is better spent on a role you actually fit.

Formatting matters only insofar as parsing matters. ATS parsers read real text in a single column best. Skip tables, text boxes, images of text, headers with critical contact info, and multi-column craziness. Use standard section names — Experience, Skills, Education — because parsers map them to database fields. A clean PDF from a text-based editor parses fine in every major system.

On keywords: the goal is not density, it's coverage of genuine skills. Read the job description, list the concrete skills it names, and make sure every one you truly have appears in your resume — in the same words the JD uses. If the JD says "PostgreSQL" and you wrote "relational databases," say PostgreSQL. Never add skills you don't have; interviews expose that in minutes, and some offer letters now include clauses letting employers rescind for misrepresentation.

Then optimize for the human scan. The recruiter's first pass looks at your current title, current company, and the top third of page one. Put your strongest, most JD-relevant evidence there: a summary that mirrors the role's language and two or three quantified wins. Numbers survive skimming; adjectives don't.

Finally, verify instead of guessing. Run your resume against the actual job description and look at the overlap and the gaps. That's exactly what TrueFit does inside JobStraight — real skill matching plus knockout detection, free, in your browser. It won't hand you a comforting fake percentage; it tells you what a recruiter would actually flag, so you can fix it before you apply.

It helps to know precisely what happens to your file after you hit submit. The ATS parses your document into structured fields — name, contact, employment history with dates, education, skills. Recruiters then search and filter that database. If parsing mangles your dates, your five years of experience may simply not exist as a filterable value, and you'll be invisible to a search for "3+ years" even though you qualify. This is the real cost of creative formatting: not rejection, but invisibility.

The six-to-eight-second human scan has a predictable anatomy. Recruiters look, in order, at your most recent job title, the company, the dates, and then whatever sits in the top third of page one. They are answering one question: is this person plausibly in the right ballpark? Everything else on the page is only read if that first answer is yes. This is why a professional summary that mirrors the role's own language outperforms a generic objective, and why burying your best achievement on page two wastes it.

Tailoring does not mean rewriting your resume from scratch for every application — that is unsustainable and unnecessary. A workable system is one master resume containing every bullet you might ever use, then a per-application pass where you reorder, cut, and adjust wording. Fifteen minutes of tailoring on a role you genuinely fit beats an hour spread across six roles you don't. Reordering alone is powerful: moving the two most relevant bullets to the top of your most recent role changes what the six-second scan lands on.

Employment gaps deserve a direct approach rather than a clever one. Functional resumes that hide chronology are widely disliked by recruiters and parse badly. Instead, keep the reverse-chronological structure and label the gap plainly — caregiving, study, redundancy, health, a failed venture. A stated gap is a non-event; an unexplained one invites the worst assumption. If you did anything relevant during it, list it as you would any role.

A 2026-specific caution: AI-written resumes have become easy to spot, and recruiters are increasingly hostile to them. The tells are consistent — inflated adjectives with no numbers, identical phrasing across every bullet, and vague claims like "spearheaded transformative initiatives." Using AI to draft is fine; shipping raw AI output is not. Rewrite anything that sounds like it could describe anyone, and make sure every bullet contains something only you could say: a real system, a real number, a real constraint.

Here is the checklist worth running before you submit. Do I actually clear the knockouts? Does the file parse as plain single-column text with standard headings? Does every genuine skill the JD names appear in my resume in the JD's own words? Are my two strongest quantified wins in the top third of page one? Have I removed anything that could describe any candidate in my field? Have I checked the dates parse correctly? Six checks, about ten minutes, and they address the things that genuinely decide the outcome — rather than the keyword-density myth that most advice still repeats.

It also helps to know that "the ATS" is not one system. Workday, Greenhouse, Lever, SmartRecruiters, iCIMS and Taleo behave differently, and the differences are practical rather than theoretical. Workday is notorious for long application forms that ask you to re-enter your entire history manually, and for parsing that frequently mangles dates — budget real time for it and check every field it auto-filled. Greenhouse and Lever are lighter, generally parse well, and are the systems where a clean PDF genuinely just works. Taleo is the oldest and the least forgiving of unusual formatting. If you find yourself repeatedly re-typing your history, that's Workday, not a sign you did something wrong.

The application form itself deserves more attention than most candidates give it, because it is often weighted more heavily than the resume. The fields you type into become structured, searchable, filterable data — while your resume is an attachment a human may or may not open. If the form asks for your years of experience with a specific technology, that answer is what gets filtered on. Fill every optional field that lets you demonstrate a genuine match, and never leave the free-text "why are you interested" box empty; it is frequently the only thing a busy recruiter reads besides your title.

Cover letters are worth a clear-eyed assessment rather than a blanket rule. Most are never read. But when a role explicitly requests one, or when your application needs explaining — a career change, a gap, a relocation, a non-obvious fit — the letter is the only place you can control the narrative. The version that works is short: three short paragraphs naming the specific problem you'd be solving, the closest thing you've actually done, and why this company. A generic letter is worse than none, because it signals volume-applying.

The uncomfortable truth is that a referral bypasses most of this. Referred candidates are interviewed at several times the rate of cold applicants at most companies, because a referral is a trust signal that no amount of keyword optimisation can replicate. This does not mean cold applications are pointless — plenty of people are hired from them — but it does mean that if you can spend twenty minutes finding one person at the company to have a genuine conversation with, that twenty minutes is worth more than another two hours of resume tweaking.

After you apply, the highest-value action is a single well-timed follow-up. Wait seven to ten days, then send one short message to the recruiter or hiring manager reiterating specific interest and adding something relevant — not "just checking in." One follow-up is professional and occasionally decisive; three is a reason to stop considering you. If you get no response after that, let it go and put the energy into the next application. Silence in modern hiring is usually operational chaos rather than a considered rejection, and reading meaning into it costs you more than it tells you.

Finally, when you do get rejected, ask once for feedback — briefly, without arguing. Most companies won't give it, but the minority who do hand you information worth more than any guide. Over a search, three or four honest pieces of feedback will tell you whether your problem is targeting, your resume, or your interviewing, and those three problems have completely different solutions. Guessing which one you have is the single most common reason a job search stalls for months without improving.

Frequently asked questions

Sources

Every figure cited above links to its origin. Where a widely-repeated statistic has no study behind it, we say so rather than repeat it.

1. Enhancv — "Does the ATS Reject Your Resume? 25 Recruiters Explain What Really Happens" (Sep–Oct 2025 study)
2. ITBrief — "Study reveals ATS rarely auto-rejects CVs, debunks 75% myth"
3. Uncharted Career — the "75% auto-rejected" myth traced to its source (Preptel, 2012)