If you take one thing from this site, take this: a two-column resume is the most common reason a qualified candidate never appears in a recruiter's search results. Not a weak summary, not the wrong font, not the length. The columns.
What a PDF actually stores
There are no columns in a PDF file. There are no paragraphs either, and no headings, and no reading order. What a PDF stores is a list of drawing instructions: put these characters at this coordinate, in this font, at this size.
The column you see is not a structure. It is an arrangement of positions that your eye interprets as a column. Nothing in the file records that the left side should be read before the right, because nothing in the format can express that.
So the parser guesses, and guesses reasonably
Faced with positions and no structure, a parser does the sensible thing: it groups text that shares a vertical position into a line, and reads each line left to right. For a normal document that is exactly correct.
Now consider a two-column resume. The left column and the right column sit at the same vertical positions, that is what makes them look like neat parallel columns. So the parser groups them together. Here is a real layout:
EXPERIENCE SKILLS Senior Analyst, Northwind SQL and query optimisation Mar 2021 – Jun 2024 Python for data pipelines Led the reporting migration Tableau dashboard design Built dashboards for 40 staff Financial modelling Analyst, Contoso plc Revenue forecasting Sep 2018 – Feb 2021 Stakeholder management
And here is what comes out:
EXPERIENCE SKILLS Senior Analyst, Northwind SQL and query optimisation Mar 2021 – Jun 2024 Python for data pipelines Led the reporting migration Tableau dashboard design Built dashboards for 40 staff Financial modelling Analyst, Contoso plc Revenue forecasting Sep 2018 – Feb 2021 Stakeholder management
Read that as a machine would. "Senior Analyst, Northwind SQL and query optimisation" is now a single field value. Your job title has been fused to an unrelated skill. Your employment dates are attached to "Python for data pipelines". The structure that made you a legible candidate is gone.
Why nobody tells you
Three reasons, and they compound.
You cannot see it. Your resume renders perfectly on your screen and prints perfectly. There is no warning, no error, no bounce. Everything about the experience tells you the document is fine.
The failure looks like ordinary rejection. You do not hear back. That happens to everyone constantly, for every reason, so it carries no information. You conclude the market is tough.
Templates encourage it. The two-column layout is genuinely good design, it uses the page efficiently and separates the scannable sidebar from the narrative. Nearly every paid resume template sells it, because it looks better in a thumbnail. The people selling templates are optimising for you buying the template.
"But my two-column resume got me interviews"
Very possibly, and this is worth taking seriously rather than dismissing.
Not every application goes through a parser. Direct email to a hiring manager, a referral, a small company, a recruiter who opens the PDF and reads it, in all of those, the layout is fine and may well help. Parsers also vary, and some modern ones do detect columns; Workday and Greenhouse handle them better than they used to.
The problem is that you cannot tell which applications went through which route. The ones where the layout cost you are precisely the ones you never heard about. Success with a two-column resume is evidence that it sometimes works, not evidence that it never fails, and the failures are invisible by construction.
How to check yours in thirty seconds
You do not need a tool for the rough version. Open your resume in any PDF reader, select all the text (Ctrl+A, or ⌘+A), copy it, and paste it into a plain text editor.
What you see is roughly what a parser sees. If the columns are interleaved, you have the problem. If it reads in order, you probably do not.
For the precise version, including which parts scramble and what else is broken, the ATS checker reconstructs both readings and shows them side by side.
The fix, and the compromise if you hate it
The fix is a single-column layout. One column, full width, top to bottom. It is less interesting to look at and it is what parses reliably everywhere.
If you genuinely cannot bear it, the honest compromise is two versions of the same resume: a single-column PDF for anything submitted through an application portal, and your designed version for direct email, your own website, and moments when you know a person is opening it. That is more work, and it is the only approach that gets you both things.
What does not work is a two-column resume with the columns "close together", or built with text boxes instead of a table, or exported differently. The parser is reading positions. If the positions form columns, it reads across them.
What to do with the space instead
The usual objection is that a single column wastes the page. It does not have to.
A skills line reads perfectly well as a single row: SQL · Python · Tableau · financial modelling · forecasting. Dates sit at the end of the role line rather than in a sidebar. A short profile paragraph runs the full width and is easier to read for it. Most of what the sidebar was doing can be done inline, and the resume usually gets shorter rather than longer.
The formatting guide covers the rest of the layout rules, and the guide to what an ATS actually does explains where the rest of the folklore comes from.