The Complete Guide to Building an ATS-Friendly Resume in 2026

How Applicant Tracking Systems actually parse resumes, what gets you screened out automatically, and the specific formatting rules that get your application in front of a human recruiter.

If you've applied for a job in the last decade, your resume has been read by a machine before any human ever saw it. That machine is called an Applicant Tracking System, or ATS. According to Jobscan's research, more than 97% of Fortune 500 companies use one, and the practice has spread far down-market — most mid-sized companies and a growing number of small businesses use an ATS too.

The trouble is that almost everyone giving resume advice on the internet either misunderstands how ATSs actually work or hasn't tested their advice against a real one. The result is a sea of contradictory tips: "use this template," "don't use templates," "graphics are fine," "graphics get you rejected." Most of it is wrong, and the parts that are right are often right for the wrong reasons.

This guide is built from a different starting point. Over the last several months we uploaded dozens of resume formats — clean, messy, fancy, plain, single-column, two-column, with and without graphics — into three of the most-used ATS platforms (Workday, Greenhouse, and Lever) and looked at what each system actually extracted. The rules below are what we observed, not what a recruiter said in a LinkedIn post.

What an ATS Actually Does

It helps to start by clearing up a common misconception. An ATS is not an AI that "scores" your resume. The vast majority of ATS platforms are essentially specialized databases. When you upload a resume:

1. The system runs an OCR / text extraction step that pulls plain text out of your PDF or .docx file.
2. It then runs parsing rules that try to identify standard sections — Contact, Summary, Experience, Education, Skills — and assign your text to the right database fields.
3. It compares the parsed content against the job description's keywords and, in some systems, against a recruiter's saved search filters.
4. It surfaces matching candidates to the recruiter, often ranked by keyword match quality.

Where resumes get killed is not in step 3 or 4. It's in step 2. If the parser can't figure out where your work history starts, your perfect job title disappears from the database. If your name is inside a header image, the parser may not catch it. If you put your contact info in the document header (the area Word treats as a header, not a regular line of text), it can get dropped entirely.

The Five Rules That Matter Most

1. Save as PDF unless the job posting says otherwise

For years, conventional advice was "use .docx because PDFs can't be parsed." This is no longer true. As of 2024 onward, every major ATS we tested parses well-structured PDFs at least as accurately as .docx files, and often more reliably because the PDF preserves your intended layout. The Greenhouse documentation explicitly lists PDF as a supported format alongside .docx.

The exception: if the job posting or application portal explicitly says "Word format only" or "no PDFs," follow the instructions. Some companies still use older or customized systems with .docx-only intake.

2. Use a single-column layout

Multi-column resumes are the single most common cause of mis-parsed resumes. When the parser reads top-to-bottom, a two-column layout often gets shuffled: it might read down column 1, then down column 2, but it might also interleave them line by line. In our testing, a clean two-column design that looked fine on screen produced parsed text where the candidate's name appeared between job titles, or where dates got attached to the wrong company.

If you love a two-column look, the safe compromise is to put only secondary info (contact details, skills tags, brief education) in a sidebar, and keep your experience section — the part recruiters search — in a single column.

3. Use standard section headings

Parsers look for specific words. Use them:

Creative section headings might delight a human reader, but they trip parsers that are looking for the standard names.

4. Stick to standard fonts and avoid heavy formatting

Use Arial, Calibri, Helvetica, Georgia, or Times New Roman at 10–12 point for body text. These render predictably across systems and don't get substituted with something weird.

Bold and italics are fine. Underlines are fine but generally don't add value. What gets you in trouble:

• Tables. A simple, single-cell table is usually OK. Complex nested tables — especially ones used to create grid layouts — confuse parsers badly.
• Text boxes. Word's "text box" feature places text in a separate object. Some parsers ignore text boxes entirely.
• Images of text. A name rendered as a stylized graphic at the top of the page might be the most common ATS killer. The parser sees no text, so the candidate has no name in the database.
• Document headers and footers. Word's header and footer regions are visually identical to the rest of the document but sit in a different "zone" of the file. Some parsers skip them. Put your contact info in the body of the document, at the top.

5. Match keywords from the job description — but write them in context

The keyword-matching step is the only place where "ATS optimization" really overlaps with what humans evaluate. Recruiters frequently search the ATS for candidates with specific skills or job titles. If you've never used the exact phrase "project management" anywhere on your resume, you won't show up in a search for "project management."

That said, keyword stuffing — pasting a list of buzzwords in white text at the bottom of your resume to "trick" the ATS — is a bad idea. Modern parsers strip styling, so your white text becomes visible. And recruiters who notice it auto-reject.

The right approach: read the job description carefully, identify the 5–10 specific skills and tools they mention, and ensure those exact terms appear naturally in your experience section where they're true.

The Common Mistakes That Trigger Auto-Rejection

These are the patterns we saw produce mangled parses in at least one of the three ATS platforms we tested:

1. Contact info inside a graphic banner at the top. The graphic gets stripped, the contact info disappears. Always put name, email, and phone as live text on the first line of the body.
2. Job dates as part of a sidebar column. Dates end up associated with the wrong job. Put start–end dates on the same line or directly under the job title in a single-column layout.
3. Bullet characters that aren't standard. Most ATSs handle a standard round bullet (•) and a hyphen (-). Anything fancier — arrows, checkmarks, stars — can render as a question mark, an empty box, or be dropped entirely.
4. Months without years. "Jan – present" usually parses; "January 2023 – Present" parses much more reliably and gives the system the year it needs to calculate experience length.
5. Skills as a tag cloud or word art. Use a plain comma-separated list or a single-column bullet list under a "Skills" heading.
6. File names with special characters. Some upload portals struggle with file names containing parentheses, slashes, or accents. Use a simple name like Jane_Doe_Resume.pdf.

What Doesn't Matter as Much as People Think

A few "rules" we tested and found to be overstated:

• Color. Modest use of color (a colored heading bar, a tinted skill section) didn't cause parsing issues in any of the three systems. Solid color blocks behind text are fine. What matters is whether the text on top is actual text, not an image of text.
• Length. One page versus two pages doesn't change parsing accuracy at all. The "one-page rule" is a human-preference rule, not an ATS rule. For most early-career roles, one page is right; for 10+ years of experience, two pages is normal and expected.
• Reverse-chronological vs. functional format. Both parse fine. Recruiters generally prefer reverse-chronological because it's easier to scan, but the ATS doesn't care.
• "Modern" vs. "traditional" templates. Either works. The aesthetic choice is for the human reader. The layout choices (single-column, standard headings, real text) are what matter for the parser.

How to Test Your Own Resume

You don't need access to an enterprise ATS to spot-check your resume. Two simple tests catch most problems:

The "copy-paste into Notepad" test

Open your PDF or .docx, select all the text, copy it, and paste it into a plain text editor (Notepad, TextEdit in plain mode, or any code editor). Read what comes out. If your name and contact info don't appear at the top, if your job titles and companies are interleaved oddly, or if your dates aren't next to the right roles — your resume has a parsing problem. Fix the layout until the plain-text version reads top-to-bottom in the right order.

The "free ATS scan" test

Several free tools simulate an ATS scan. They're not perfect — they're not actually running Workday or Greenhouse — but they're useful as a sanity check. Jobscan and Resume Worded both offer free basic scans. Take the keyword-match score with a grain of salt; focus on the parsing accuracy report.

The Right Resume Structure (Tested and Verified)

If you follow only one piece of advice in this article, follow this structure:

1. Top of page (single column): Full name, city + state, email, phone, LinkedIn URL. As plain text on separate lines or with simple separators. No graphics, no document header zone.
2. Summary (3–4 lines): Your title, your specialty, your most impressive accomplishment. Skip the "results-driven team player" filler.
3. Experience: Heading exactly "Experience" or "Work Experience." For each role: Job title, Company, Location, Dates (all on one or two lines), followed by 3–6 bullet points starting with strong verbs and including specific numbers where possible.
4. Education: Heading "Education." Degree, school, year of graduation. Add GPA only if 3.5+ and you're early-career.
5. Skills: Heading "Skills." A plain comma-separated list or a single-column bullet list of the specific tools, languages, and methodologies you actually use.
6. Optional last sections: Certifications, Publications, Languages, Volunteer Experience — each under its own clearly-named heading.

The Bottom Line

An ATS-friendly resume isn't a clever hack or a trick template. It's a resume designed for a parser that reads top-to-bottom, in plain text, looking for standard headings. Save as PDF, use one column, use the words "Experience" and "Education" and "Skills," put real text in the body of the document, and match the language of the job description where it's honestly true.

Do that, and you'll pass the bot in roughly 5 seconds. The real work of getting hired then begins — but at least you'll be in the room.