Can an ATS Read a Two-Column Resume? We Tested Our Own Templates

What's documented

Greenhouse's support page on unsuccessful parses names columned layouts, tables, and text in headers, footers or text boxes as things that can stop a resume parsing properly (Greenhouse Support). The reason is reading order. A parser gets text out of the file in some order, and with columns, the order has to be guessed.

How we tested

• Resume: one fictional analyst resume ("Alex Morgan", sample content, one page).
• Templates: Redvica's single-column "Modern ATS" and its "Two-column Modern" with a sidebar, both exported as PDF by the resume builder's own PDF generator.
• Extractors: pdfminer.six (default settings), pypdf, pdftotext (default) and pdftotext in layout mode, which are common open-source tools for reading text out of PDFs.
• What we looked at: where the name and contact details land, whether text from the two columns mixes, and where each job's dates end up.

Result 1: the name no longer comes first

In the two-column file, three of the four extractors read the whole sidebar (contact details, skills, certifications) first and the name only afterwards. A system that assumes the name is at the top could mis-assign it. The text blocks themselves stayed intact, though: no extractor split a sentence between columns in its default mode.

Result 2: layout mode mixes the columns line by line

pdftotext's layout mode keeps text where it sits on the page. In a two-column file, that means one output line holds text from both columns:

A parser that processes line by line would see a phone number joined to a job title, or a skill joined to the end of a sentence. That's the failure the documentation warns about.

Result 3: dates drifted in both layouts

The most surprising result was not about columns. In both templates, each job's date range is right-aligned on the same row as the job title. Whether the extractor kept it with its job depended on the tool:

So even a clean single-column file can lose the link between a role and its dates in some extractors, because right-aligned dates sit far from the title. Employment dates are a field recruiters filter on. A practical fix for anyone writing their own file is to put the date on the same line right after the title, not at the far edge, and to check the pasted text (see the copy-and-paste test).

What this means for your resume

• For upload forms and online applications, use a single column. It removes the reading-order guess that columns create.
• Use two-column designs when a person will read the file: emailing a recruiter, networking, a portfolio. That's where they look their best.
• Whatever the layout, test the text order. Select all, paste into a plain text editor and read it.
• Keep contact details and your name where a parser expects them: at the top of the main text, not only in a sidebar.

Redvica's resume builder has five single-column and five two-column templates; see the templates page. We don't claim the two-column ones are safe for upload forms. We built them for readability by people.

Limits of this test

• One fictional resume, two templates and four open-source extractors. Results can differ with other layouts, tools and file creators.
• Real applicant tracking systems may handle columns better or worse than these tools. Employers also configure them differently.
• We tested our own templates, so the results describe our PDF output, not other builders' files.