Resumes
ATS Resume Format
The layouts, fonts, and file types that parse cleanly in modern ATS platforms—and the two-reader test that catches format failures before they cost you interviews.
Here is the part most resume advice skips: in the majority of applicant tracking systems, the recruiter never opens the PDF you spent hours designing. They work from the parsed version—the plain fields the software extracted. Name, titles, employers, dates, skills. If the parser mangles those fields, your resume effectively arrived blank, and nobody will tell you.
That makes format the cheapest fix in your entire job search. It's not a per-application task like keywords or tailoring. You fix it once, and every application after that benefits.
This guide is part of the Ultimate ATS Resume Guide. Use the pillar for the full map; use this page to get your format right in one sitting.
Your resume gets read twice, and the first reader is software
Over 90% of employers use an ATS or recruiting management system to filter and rank candidates, according to the Harvard Business School and Accenture report Hidden Workers: Untapped Talent (2021). The same report found that 88% of employers believe qualified, high-skill candidates get vetted out of the process because they don't match the exact criteria in the job description.
Formatting failures are a big, invisible slice of that. A candidate who has the exact skills the role needs, listed in a sidebar the parser couldn't read, looks identical to a candidate who doesn't have them.
The second reader is a human under time pressure. Eye-tracking research from Ladders (2018) put the initial recruiter scan at 7.4 seconds. Both readers need to succeed, in that order—a resume that only works for one of them fails.
What actually breaks parsers
Parsers read text in document order, top to bottom. Everything that breaks parsing breaks it by disrupting that order or hiding text from extraction:
| Risky choice | What goes wrong | Parse-safe alternative |
|---|---|---|
| Two-column or sidebar layout | Columns get interleaved in reading order | Single column, top to bottom |
| Contact info in header/footer | Many parsers skip those regions | Contact block in the body, top of page one |
| Tables and text boxes | Cell contents extracted out of sequence, or not at all | Plain text with simple bullets |
| Icons, charts, skill bars | Invisible to text extraction | Words and numbers |
| Creative headings ("My Journey") | Section mapping fails; experience lands in the wrong field | Experience, Education, Skills, Certifications |
| Scanned or design-tool PDF | No selectable text to extract | PDF exported from Word/Google Docs, or DOCX |
Here's what interleaving looks like in practice. A software engineer applying for a React role submits a two-column template: skills in a left sidebar, experience on the right. The parser reads across, and the extracted experience section comes out as:
"Senior Frontend Engineer React TypeScript Acme Corp Node.js GraphQL Led migration of Jest 40-screen dashboard..."
The recruiter searching the ATS for "React" might still find this candidate. But the recruiter reading the record sees word salad where the work history should be, and moves to the next of two hundred applicants. The resume was never rejected. It was just never legible.
The fix takes one afternoon
Start from one column. Top to bottom, 0.5–1 inch margins, a standard font (Calibri, Arial, Georgia, Helvetica) at 10–12pt. Boring is a feature—every second the reader spends decoding your layout is a second not spent on your experience.
Use the headings parsers are trained on. Summary, Experience or Work Experience, Education, Skills, Certifications. Parsers map your content into fields by matching these headings. "Where I've Made an Impact" breaks the mapping, and your work history lands in the wrong field or nowhere.
Keep every role line in the same pattern. Title, company, location, dates—same order, every role. Parsers infer the pattern from repetition; inconsistency makes them guess, and they guess wrong.
Write dates one way, everywhere. Mar 2023 – Present or 03/2023 – Present, but not both. Mixed date formats are the most common cause of ATS records showing phantom employment gaps.
Put contact details in the body. Name, email, phone, city, LinkedIn—at the top of page one, not in the header. This one is worth repeating because the failure mode is so silent: an application filed with no contact info attached doesn't bounce. It just never gets a reply.
Export a file with selectable text. A PDF from Word or Google Docs is fine. A Canva export, a Photoshop export, or a scan is not. The test takes two seconds: if you can't select the text with your cursor, neither can the parser.
Format checklist — done once, benefits every application
- Single column, standard font, 10–12pt, 0.5–1 inch margins
- Standard section headings (Experience, Education, Skills)
- Same role-line pattern for every job: title, company, location, dates
- One date format used consistently throughout
- Contact block in the body at the top—nothing important in headers or footers
- No tables, text boxes, columns, icons, or skill bars
- Text is selectable in the exported file
- File named FirstLast-Role.pdf, not resume_final_v7.pdf
The two-reader test
Before you spend a minute on keywords, run both passes. This is the check we built into Kyrolane's own resume workflow, and it catches the large majority of format problems before they cost anyone an application:
Machine pass. Select all the text in your exported file and paste it into a plain-text editor. Read the result top to bottom. Sections in order? Dates attached to the right roles? Nothing from a sidebar spliced mid-sentence? If your own paste is scrambled, the ATS output is worse.
Human pass. Hand the formatted version to a friend for seven seconds—roughly what the Ladders research says you get. Then ask them three questions: What role is this person going for? What's their current title? Name one impressive thing they did. Three answers means the layout works. Blank stares mean the format is hiding your best material, even if it parses perfectly.
What we learned building a resume parser
One thing surprised us when we built resume analysis into Kyrolane: the failures almost never look like failures. A resume with a table-based skills section doesn't produce an error. It produces a plausible-looking record with an empty skills field, and everything downstream—matching, ranking, recruiter search—operates on that emptiness with full confidence.
That's why we put a parse check at the front of the workflow instead of treating format as a cosmetic afterthought, and why a human reviews every suggested fix before anything changes. The parser tells you what the machines see; you decide what to do about it.
Related reading
- Ultimate ATS Resume Guide — the full system, from format to tracking
- Resume Keywords Guide — what to do after your format parses cleanly
- How Applicant Tracking Systems Rank Candidates — what happens to your parsed data next
- Get started with Kyrolane