By Peter Miller · Published March 24, 2026 · Last updated: September 16, 2026
11 min read · Pretty PDFs still fail plain-text imports
Your resume looks sharp on your laptop. Workday shows your skills where your job title should be. You're not failing a secret keyword test. You're failing reading order. Parsers treat the file like a tape: they pull characters in sequence. Columns and sidebars split that tape in the middle, and you can't keyword your way past a scrambled import.
If you've been told to "make it stand out," that advice wasn't wrong for humans. It was incomplete for parsers. Standing out with layout often means standing in the wrong search bucket. We'll fix that without making your file ugly to recruiters who do open the PDF.
Before you rewrite bullets, check your resume for free and confirm title lines imported above employer names. Ten minutes on layout beats an hour on synonyms when the ATS never attached dates to the right role.
Below you'll see why single column wins for US corporate ATS, the narrow exceptions where design still works, before-and-after layout fixes, a copy-paste skeleton, and what to do tonight if your preview already looks wrong. You don't need a designer. You need a boring file that imports cleanly.
I've opened thousands of parser previews next to the PDF candidates swore was fine. The pattern is always layout before language. Fix the column, then argue about wording.
Quick Wins
- Delete columns and sidebars; move skills under Experience or a plain Skills header.
- Put job title on the line above employer; keep dates on the same line when possible.
- Run the Notepad test on your export before you tailor keywords.
Why the safest resume layout for US ATS drops columns
US employers run Greenhouse, Workday, Lever, Taleo, and iCIMS more often than boutique parsers. Those systems were built for forms, not magazine spreads. They reward vertical stacks: header, summary, experience block, education, skills. Anything that fights vertical flow becomes noise in the import.
Columns scramble association. Your left column might list employers while your right column lists tools. The parser reads across and pairs Java with the wrong company. Recruiters search title plus employer plus years. Scrambled pairs drop you out of default lists even when the PDF looks logical to you.
Sidebars steal title lines. Contact blocks trapped in colored rails sometimes import as body text while certifications jump into the title field. I've seen Registered Nurse land under Skills while the Title row stays blank. That's a layout failure, not a wording failure.
Humans skim in an F-pattern on screen, but parsers do not. They do not know your logo is decorative. They do not know the gray box is a skills cloud. They read characters until they stop. Single column removes the guesswork.
Keyword stuffing cannot fix scrambled order. If bullet three imports before bullet one, the posting's top skill might sit on the wrong job. Fix structure, then tailor language.
Mobile previews make the problem worse. Recruiters open files on phones during meetings. A sidebar that collapses oddly on mobile often means the desktop PDF never matched the import either. One column reads the same in every viewport.
International candidates sometimes bring two-page EU layouts with photo boxes. US corporate ATS uploads should drop photos and move to letter-size single column even if your home market expects a portrait. The portal does not care about tradition; it cares about field order.
Contract and consulting histories tempt people into project grids. List clients vertically with title lines and dates like any other employer. Grids look organized to you and random to a parser.
When you inherit an old Word template from a career center, strip its columns before you update bullets. Career centers optimize for print handouts at job fairs, not for Greenhouse imports at midnight.
For how normalized text behaves inside matchers, read how ATS normalizes resume text and why it matters . Clean vertical order survives normalization better than nested tables.
For title lines that vanish in sidebars, see how recruiters interpret job titles . Layout and headline string fail together more often than candidates admit.
What the safe layout looks like in practice
Think plain document, not portfolio. One margin, one font family, bold section labels, hyphen bullets. The goal is identical reading order in Word, PDF export, and the portal preview.
Before and after: two-column contact header
Before: Name and title in a left cell, email and phone in a right cell inside a table row.
After: Name centered or left-aligned on line one, title on line two, email | phone | city on line three, all plain text with no table.
Before and after: skills sidebar
Before: Skills listed in a right column beside experience bullets.
After: Skills section after Experience with comma-separated tools, plus each top skill repeated once inside a relevant bullet for searchability.
Before and after: dates floating in margin
Before: Dates right-aligned in a separate text box per job.
After: Marketing Manager | Acme Corp | Jan 2019 – Dec 2022 on one line, bullets directly underneath.
Before and after: graphic skill meters
Before: Five-star Excel rating and icon rows for tools.
After: Built rolling forecast model in Excel used in weekly CFO review, with Skills line listing Excel, SQL, and Looker in plain text.
Copy-paste single-column skeleton
YOUR NAME Target Job Title (optional, matches posting) [email protected] | phone | City, ST | linkedin.com/in/you SUMMARY Two lines max. Name segment, years, and top tool from the posting in plain English. EXPERIENCE Job Title | Employer Name | City, ST | Month Year – Month Year - Bullet with verb, object, and metric - Bullet tying posting keyword to proof - Bullet with scope (team size, budget, volume) Job Title | Employer Name | Month Year – Month Year - Bullet - Bullet EDUCATION Degree | School | Graduation Year SKILLS Tool, tool, tool (plain comma list, no icons) CERTIFICATIONS (optional) Credential | Issuer | Year
Copy-paste the skeleton into Word, replace placeholders, and export a simple PDF only if the posting allows PDF. Keep section headers as SUMMARY and EXPERIENCE. Creative headers like My Journey confuse parsers trained on US defaults.
Fonts: Arial, Calibri, Georgia, or Times New Roman at 10–12 pt body. Fancy thin fonts break in older imports. Consistent Month Year dates across every role.
Edge case: federal resume builders with rigid fields. Follow their boxes even if it feels ugly. Custom columns inside those uploads still break downstream.
Edge case: creative portfolios for design roles. Keep the ATS upload boring and link a portfolio site in the header line. Do not merge portfolio grids into the same file you upload to Workday.
Line spacing matters more than people think. Single spacing with a blank line between jobs beats tight blocks that parsers read as one paragraph. You are not trying to win a page-count contest with the posting; you are trying to win correct field mapping.
Bullet characters should stay boring. Hyphens or simple round bullets survive. Fancy Wingdings-style symbols sometimes import as question marks, which makes your metrics look like typos.
If you use a summary, keep it two lines and repeat the posting title string once. Long summary paragraphs push first job experience below the fold in some previews. Recruiters still read summaries, but parsers anchor on experience headers first.
When you need a faster sanity check, read how to tell if your resume is ATS friendly in 60 seconds . It walks the Notepad test step by step.
Exceptions and layout traps that still fail parsers
Some postings allow designed PDFs for brand-facing roles. Most corporate reqs still want the boring file. When in doubt, choose boring for the upload and pretty for the human follow-up email after a screen.
Before: You keep a two-column template because it fit on one page for printing.
After: You allow a second page before you allow a sidebar. Parsers handle length better than grids.
Tables for alignment look like columns to parsers. Using invisible borders still counts as a table. Use tabs sparingly; spaces and pipes are safer.
Headers and footers with your name or page numbers often drop on import. Keep contact info in the body top. Recruiters rarely print anyway.
Text boxes in Canva or Illustrator exports frequently flatten wrong. Build in Word or Google Docs, then export. Design tools are for slides, not Workday uploads.
Multiple columns in Google Docs resume templates are popular and risky. Strip them even if the thumbnail looks professional.
Page breaks mid-job are rare but painful. If a job header sits at the bottom of page one and bullets start page two without repeating context, some parsers attach bullets to the previous employer. Keep each role's header plus at least one bullet on the same page when you can.
Hyperlinks in headers sometimes survive; sometimes they vanish. Spell out linkedin.com/in/you in plain text in the contact block instead of hiding URLs behind icons.
For skills-heavy clusters, read resume skills section how to make it readable for ATS and humans . Skills belong in plain lists, not floating widgets.
Test the layout before the next application
Upload your rebuilt file to HireFlow's free ATS resume checker and compare extracted order to your on-screen PDF. If titles and employers swap places, stay on layout until the preview matches the skeleton above.
Build an ATS resume in a single-column template if you are starting from a design file you cannot salvage. Faster to rebuild once than to lose weeks on silent rejections.
When a posting asks for a letter, use the cover letter generator after your layout passes preview. Letters cannot fix a scrambled experience block in the upload.
Score your job match only after structure is clean. Keyword gaps matter after parsers can read bullet one on the right job.
Frequently asked questions
Most parsers read left to right across the full page width, not down each column. Skills in a right rail can land above job titles. Dates from a sidebar can attach to the wrong employer. Recruiters then search a scrambled profile even when your PDF looks fine on screen. A single column keeps title, employer, and dates on one vertical path.
No. Tables often break the same way columns do. Cells import out of reading order or drop entirely in older Taleo and iCIMS builds. Use line breaks and bold section headers instead of grid layouts. If you need two pieces of info on one line, separate them with a pipe character on plain text, not a table cell.
Follow the posting. When the portal accepts both, Word usually survives picky parsers better because text stays editable. Simple PDFs exported from Word are fine when the posting asks for PDF only. Fancy design PDFs with text boxes fail more often. Test the portal preview after upload instead of guessing by file type.
Skip icons, rating bars, and chart graphics. They look modern on LinkedIn but often import as blank squares or push text out of order. List skills as plain words in a single column block or weave them into experience bullets. Humans still need readable proof lines; parsers need plain characters in order.
Upload to the employer portal and open the applicant preview if available. Copy the imported text into Notepad on Windows or TextEdit in plain text mode on Mac. If job titles sit above employers and dates follow in order, you passed. If skills jumped to the header, rebuild as one column before you apply again.