Does ATS Read Resume Columns and Tables? What Actually Happens

You spent hours building a clean two-column resume in Word or Canva. It looks polished. The problem is that what looks good to a human eye can look like scrambled noise to an applicant tracking system. Before you submit that file, it’s worth understanding exactly what happens inside the parser — not just vague warnings about “ATS compatibility.”

How Major ATS Platforms Parse Resume Layout

Not all ATS platforms use the same parsing engine, and the differences matter.

Workday uses its own proprietary parser. It reads documents in a linear, top-to-bottom, left-to-right order. When it encounters a two-column layout built with a table or text boxes, it often reads across both columns simultaneously rather than finishing the left column first. The result: your job titles get merged with your skills list, and dates end up next to the wrong employer.

Greenhouse relies on third-party parsing engines (historically Sovren/Textkernel). These engines handle simple two-column layouts better than Workday’s parser, but they still struggle with nested tables, text boxes, and columns created via invisible table borders.

Lever parses the plain text it can extract from your file. Lever’s candidate profile view shows you exactly what it captured — and if you’ve ever pasted a formatted resume there and seen the output, you know how much data disappears from complex layouts.

iCIMS uses its own extraction layer. It handles standard single-column Word documents reliably, but multi-column layouts and PDF files with embedded text boxes frequently produce garbled output or blank fields in the candidate profile.

The common thread: every major ATS reads text linearly. Layout elements that force text into non-linear positions — columns, tables, text boxes, headers, footers — create parsing errors at different rates depending on the platform.

Columns: Single vs. Two-Column Test Results

Testing resume formats against actual ATS parsers produces consistent findings. A single-column resume in a standard .docx file parses cleanly across Workday, Greenhouse, Lever, and iCIMS roughly 95% of the time. A two-column layout created with a Word table drops to somewhere between 60-75% clean extraction, depending on the platform and file type.

The specific failure mode for columns: the parser reads across the page horizontally. If your left column contains “Project Manager” and your right column contains “Python, SQL, Tableau,” the parser may output “Project Manager Python, SQL, Tableau” as a single string — destroying both your job title field and your skills section.

PDF files make this worse. When you export a two-column Word document to PDF, the text layer in the PDF often encodes columns as interleaved text streams. Parsers that rely on PDF text extraction then produce output that alternates between columns line by line.

If you’re using a visually formatted template, run a quick test: copy and paste the text from your PDF into a plain text editor. If the order looks wrong, the ATS parser will likely produce the same scrambled result.

For a reliable starting point, see the ATS-friendly resume template guide — it covers which structural choices consistently pass parsing.

Tables and Text Boxes: What Gets Dropped

Tables and text boxes are the two formatting elements most likely to cause complete data loss, not just scrambling.

Text boxes in Word and Google Docs are rendered as floating objects. Most ATS parsers treat floating objects as images — they skip them entirely. If your contact information, summary, or skills section lives inside a text box, the parser may extract zero data from it. Your phone number, email, and LinkedIn URL can vanish completely.

Tables behave differently depending on how they’re used. A simple two-row, two-column table used to align a job title and date is usually parsed, though the order may be wrong. A complex skills table with multiple rows and columns is frequently extracted as a single block of merged text, making keyword matching unreliable.

Graphic elements — icons, progress bars for skills, profile photos placed in table cells — are treated as images and ignored by every major parser. A skills section that shows five filled circles to indicate proficiency level tells the ATS nothing. The word “Python” needs to appear as plain text.

Headers and footers in Word documents are another common trap. Workday and iCIMS frequently skip header and footer regions entirely. Placing your name and contact details in the document header — a common design choice — means the ATS may create a candidate profile with no name and no email address.

Header and Footer Parsing Issues

This is worth its own section because the failure is so common and so damaging.

Word’s built-in header and footer zones are separate content streams from the main document body. When an ATS extracts text, it pulls from the main body first. Some parsers never reach the header/footer stream at all. Others extract it but append it to the end of the document, so your name appears after your references section in the parsed output.

The fix is simple: move your name, phone number, email, LinkedIn URL, and location into the main document body. Use a regular paragraph at the top of the page, not a header zone. This one change eliminates a significant source of parsing failures.

For a broader look at what ATS systems actually do with your file before a human sees it, the ATS optimization guide covers the full workflow from submission to recruiter review.

Safe Formatting Rules for ATS Compliance

Here is what consistently works across the major platforms:

• File format: .docx is safer than PDF for most ATS platforms. If a job posting specifies PDF, use it — but if you have a choice, .docx parses more reliably.
• Layout: Single column. No exceptions if ATS compliance is the priority.
• Fonts: Standard fonts (Calibri, Arial, Georgia, Times New Roman) at 10-12pt. Decorative fonts sometimes convert to symbols in parsed output.
• Contact info: In the main document body, not in a Word header zone.
• Section headers: Plain text with consistent formatting. Avoid text boxes, shapes, or graphic dividers.
• Tables: Avoid them for layout purposes. A simple table for a publication list or a two-column skills list is lower risk, but still not zero risk.
• Text boxes: Remove them entirely.
• Graphics and icons: Remove them. They add no value to a parser and can interrupt text extraction.
• Bullets: Standard round bullets (•) or hyphens. Decorative bullet symbols sometimes convert to question marks or boxes.

If you’re tailoring your resume for a specific role, keeping the format clean also makes keyword optimization more effective — the parser needs to read your content before it can match your keywords to the job description. The guide on how to tailor your resume to a job description walks through that process in detail.

When Visual Formatting Still Makes Sense

ATS compliance isn’t the only consideration. If you’re submitting directly to a hiring manager via email, attending a career fair, or uploading to a portfolio site, a visually formatted resume can make a stronger impression. The practical approach: maintain two versions. Keep a clean, single-column .docx for ATS submissions and a designed PDF for direct human review.

Some roles — graphic design, marketing, UX — have hiring processes where portfolio work matters more than resume parsing. Even then, if the application goes through an ATS first, the plain version gets you past the filter before the designed version gets seen.

For a complete picture of what makes a resume work in 2026, the resume tips guide covers both ATS and human review stages together.

If you want to check how well your current resume is being parsed before submitting, TailorMyJob lets you upload your resume and see what content is being extracted — so you can catch formatting problems before they cost you an interview.