What actually makes a résumé ATS-safe
Applicant tracking systems are simpler and dumber than people fear. Here is what they read, what breaks them, and why a single column is the whole trick.
Most résumé advice about ATS is fear dressed up as expertise. The truth is calmer: an applicant tracking system is a parser. It turns your file into plain text and key fields, then a recruiter searches and filters that text. If the parser reads you cleanly, you are in the pile. If it garbles you, you are not. That is the entire game.
What the parser actually does
When you upload a PDF or DOCX, the system extracts a stream of text and tries to label parts of it: name, contact, work history, education, skills. It does this with layout heuristics. Anything that confuses the reading order or hides text inside a graphic makes the labels wrong.
- Reads top to bottom, left to right. Two columns can interleave into nonsense.
- Pulls text, not pictures. A name inside a logo or header image is invisible.
- Maps standard headings. "Work Experience" parses; a clever synonym may not.
- Keeps simple runs of text. Tables, text boxes and multi-column grids scramble order.
The things that quietly break you
None of these are about taste. They are about whether a machine can read you at all.
- Multiple columns. The single most common reason a clean-looking résumé parses badly.
- Headers and footers holding your contact details, which some parsers drop.
- Icons standing in for labels, so "email" and "phone" never appear as text.
- Tables for layout, which reorder cells unpredictably.
- Exotic fonts or text saved as an image, which can come through empty.
Why one column wins
A single-column layout with standard section headings has exactly one reading order, and it is the one you intended. There is nothing to interleave, nothing hidden in a sidebar, nothing for the parser to guess. It looks plainer than a two-column showpiece, and it is read correctly far more often. For a document whose only job is to be read by a stranger and a machine, correct beats clever.
How lapel handles it
You cannot pick an option in lapel that breaks parsing. Every template is single-column with standard headings and text-based export to PDF and DOCX. The preview is the document, so what you see is what the parser sees. You spend your effort on the words, which is the part that is actually hard.