Why Single-Column Resumes Outperform Two-Column Layouts for ATS

The Layout Decision That Silently Kills 43% of Applications

Every year, millions of software engineers lose job opportunities before a human ever reads their resume. The culprit is not a missing keyword or a gap in employment history — it is a two-column layout that an Applicant Tracking System cannot parse correctly. If you have ever applied to 200 jobs and heard nothing back, your resume's column structure may be the single highest-leverage fix available to you.

This is not a matter of aesthetic preference. It is a technical problem rooted in how document parsers extract text from PDF and DOCX files. In this guide, we will walk through exactly how ATS parsers read your resume, what happens when they encounter multi-column layouts, and why the data overwhelmingly favors single-column formatting — even when two-column designs look better to the human eye.

How ATS Parsers Actually Read Your Resume

An ATS does not see your resume the way you see it. There is no rendering engine displaying your careful typography and balanced whitespace. Instead, the system extracts raw text from your uploaded file, then runs that text through a series of classifiers that attempt to identify sections (work experience, education, skills) and map individual data points (job titles, company names, dates, degree types) into structured fields in a database.

The critical detail is this: text extraction is linear. Parsers read content in the order it appears in the underlying file structure, which for a single-column document maps cleanly to top-to-bottom, left-to-right reading order. The parser encounters your name, then your contact info, then your summary, then your experience — exactly as you intended.

Workday, Greenhouse, Lever, iCIMS, and Taleo — the five ATS platforms that collectively process over 75% of Fortune 500 job applications — all use some variant of this linear extraction approach. While their proprietary parsers differ in sophistication, none of them reliably reconstruct the visual spatial relationship between two adjacent columns of text.

Parsing Failure Rates: What the Data Shows

Jobscan's 2023 analysis of over 1 million resume submissions found that resumes using two-column layouts experienced a 43% higher rate of critical parsing errors compared to single-column resumes. Critical parsing errors are defined as failures where the ATS either misidentifies a section header, merges content from two unrelated sections, or drops content entirely.

Lever's engineering team published a blog post in 2022 discussing their parser improvements, noting that multi-column detection remained one of their most persistent challenges. Greenhouse's support documentation explicitly recommends single-column formatting for optimal parsing results. These are not fringe opinions — they are the platform vendors themselves telling you how their own software works.

We've invested significantly in handling edge cases around document layout, but the reality is that a cleanly structured single-column resume will always parse more reliably than a multi-column alternative. The physics of text extraction favor simplicity.

The CSS Float and Table Problem in DOCX Files

If you built your two-column resume in Microsoft Word or Google Docs, the columns are almost certainly implemented using either a table structure or text columns defined in the document's XML. Here is why that matters technically.

A DOCX file is a ZIP archive containing XML files. The main document body lives in word/document.xml. When you create a two-column layout using a table, the XML nests your content inside <w:tc> (table cell) elements within <w:tr> (table row) elements. Most ATS parsers read table cells sequentially — left cell first, then right cell — row by row. This means if your left column contains your skills list and your right column contains your work experience, the parser may interleave skills and job titles in ways that make the extracted text incoherent.

Text columns (the Word feature under Layout > Columns) create a different but equally problematic XML structure using <w:cols> elements. The text reflow behavior that makes columns look correct on screen has no equivalent in the flat text extraction pipeline. The parser sees one continuous stream of text and has no reliable signal for where the left column ends and the right column begins.

Why PDF Two-Column Layouts Break Parsing Order

PDF files present an even more insidious version of this problem. Unlike DOCX, a PDF does not have a semantic document structure by default. A PDF is essentially a set of drawing instructions: place this glyph at coordinates (x, y), draw this line from point A to point B. There is no inherent concept of paragraphs, columns, or reading order.

When an ATS parser extracts text from a PDF, it must reconstruct reading order from spatial coordinates. The standard approach is to sort text objects by their Y coordinate (vertical position) first, then by X coordinate (horizontal position) within the same vertical band. For a single-column document, this produces perfect results. For a two-column document, the parser must determine the column boundary — and it frequently gets this wrong.

1. The parser reads line 1 of the left column and line 1 of the right column as a single line, because they share the same Y coordinate
2. Section headers in the left column get concatenated with body text in the right column at the same vertical position
3. Bullet points from the left column's experience section merge with dates or skills from the right column
4. The parser may detect the column gap as a section break, splitting one job entry into two incomplete records
5. Tagged PDFs (PDF/UA) can provide reading order hints, but most resume-generating tools do not produce properly tagged output

Tools like LaTeX, Canva, and many online resume builders generate PDFs where the content stream order reflects the order elements were added to the design canvas — not the visual reading order. A two-column LaTeX resume using the minipage environment writes the entire left column to the content stream first, then the entire right column. Some parsers handle this correctly. Many do not. You are gambling with every application.

The 'Human Readability' Myth

The most common argument for two-column layouts is that they are easier for humans to scan. This claim deserves scrutiny, because the research does not support it as strongly as design-focused resume advice suggests.

Ladders' 2018 eye-tracking study — the most widely cited research on recruiter reading patterns — found that recruiters spend an average of 7.4 seconds on an initial resume screen. The eye-tracking heatmaps from that study show an F-shaped scanning pattern: recruiters read the top of the resume more thoroughly, then scan down the left side. This pattern maps naturally to a single-column layout. The study did not find that two-column layouts improved scan speed or information retention.

There is a secondary consideration that two-column advocates overlook: recruiter context switching. When a recruiter reviews your resume in an ATS interface, they are not looking at your beautifully formatted PDF. They are looking at the parsed data displayed in the ATS's own UI — structured fields for job title, company, dates, and bullet points. Your column layout is invisible to them. The formatting only matters if they click through to view the original document, which Greenhouse usage data suggests happens for fewer than 30% of candidates who pass the initial ATS screen.

What FAANG Recruiters Actually Say

Internal recruiters at large technology companies have been remarkably consistent on this topic in public forums, conference talks, and published interviews. The consensus is unambiguous.

I've reviewed over 10,000 resumes in my career. The candidates who use fancy two-column templates from Canva or Etsy are disproportionately the ones whose profiles show up in our system with scrambled data. I then have to decide whether to spend 3 minutes manually fixing their record or move on to the next candidate. Most of the time, I move on.

ATS-by-ATS Breakdown: How Each Platform Handles Columns

Workday

Workday's parser has improved significantly in recent years but still struggles with table-based column layouts in DOCX files. PDF parsing uses a coordinate-based text extraction that handles simple two-column layouts moderately well but fails on designs with overlapping vertical sections, sidebars with background colors, or columns of unequal width. Workday processes applications for approximately 50% of Fortune 500 companies.

Greenhouse and Lever

Both platforms are popular with startups and mid-size tech companies. Greenhouse uses a combination of internal parsing and third-party extraction services. Lever's parser is generally regarded as more modern, but both platforms' own documentation recommends simple formatting. Greenhouse's support article on resume formatting explicitly states: 'Avoid tables, columns, headers, footers, and text boxes for the best parsing results.'

iCIMS and Taleo

These legacy platforms power hiring at many large enterprises and government contractors. Their parsers are notably less sophisticated than Workday or Greenhouse. Taleo in particular is known for aggressive text flattening that destroys column relationships entirely. If you are applying to any company running Taleo — and you often cannot know in advance — a two-column resume is almost guaranteed to parse incorrectly.

Actionable Takeaways: How to Format Your Resume for Maximum Parseability

1. Use a single-column layout for all ATS-submitted resumes — no exceptions, no sidebars, no skill bars on the left margin
2. If you must use a two-column design for networking or direct emails, maintain a separate single-column version for online applications
3. Avoid tables, text boxes, and floating elements in your DOCX files — use simple paragraphs, headings, and bulleted lists
4. If submitting a PDF, test it by selecting all text (Ctrl+A) and pasting into a plain text editor — if the order is wrong there, it will be wrong in the ATS
5. Use standard section headers: 'Experience,' 'Education,' 'Skills,' 'Projects' — creative alternatives like 'Where I've Made Impact' confuse section classifiers
6. Place your name and contact information at the very top, not in a header or footer element that parsers may skip entirely
7. Use reverse chronological order within each section — parsers expect and optimize for this pattern
8. Avoid icons, images, charts, and graphical skill-level indicators — they are invisible to text extraction and waste valuable space

The One Exception: When Two Columns Are Acceptable

There is exactly one scenario where a two-column layout is defensible: when you are handing a physical copy of your resume to a human being, or attaching it to a direct email where no ATS will process it. Portfolio reviews, career fairs, and networking events are contexts where visual design legitimately matters and parseability is irrelevant.

For every other submission channel — online job applications, recruiter portals, LinkedIn Easy Apply, company career pages — single-column is the only rational choice. The asymmetry of outcomes is too stark: a well-formatted single-column resume loses you nothing with human readers and gains you everything with automated systems.

The Bottom Line

Two-column resumes are an optimization for the wrong audience. They prioritize the 30% of cases where a human views your original document over the 100% of cases where an ATS must parse it correctly. The data from Jobscan, the documentation from Greenhouse and Lever, and the consistent advice from FAANG recruiters all converge on the same conclusion: single-column layouts outperform two-column layouts for ATS compatibility by a wide margin.

Your resume's job is to get you interviews. It cannot do that job if it is sitting in a database with your job title listed as 'React Senior Engineer' because a parser merged your skills column with your experience column. Choose the format that works with every system, every time. Choose single-column.