Most failures are mechanical, not editorial
An applicant tracking system pulls text out of your file and sorts it into fields: name, contact, employers, dates, education. Most of what goes wrong there is mechanical — the file is a picture rather than text, or the layout gives no clear reading order.
Both export routes here produce real text, and the single-column layout gives a parser one order to read in.
Test your own file in ten seconds
Open the PDF, select all, copy, and paste into any plain text editor.
What appears is roughly what a parser sees. If nothing selects — if the page behaves like an image — no tracking system can read it, and neither can anyone searching their own database for your name later. That happens with scanned resumes and with files exported as images. Nothing else on this page matters if it does.
Read the pasted text for two things: is your name at the top, and are the employers and dates still attached to the right entries.
What the exports do about it
One-click download generates the PDF directly, as text. Section headings are set without letter spacing, on purpose: tracked-out headings extract character by character, so "EXPERIENCE" arrives as E X P E R I E N C E and the parser stops recognising the section. The spacing looks nice and costs you the section heading, so we don't use it.
Printing hands the job to your browser, which also writes text, not a screenshot. Two useful side effects: two-column layouts straighten into a single column on paper, so the reading order becomes linear, and the editor's own buttons and navigation never reach the file.
Logos are images placed beside an entry. A parser ignores them; the company or school name sitting next to the logo is ordinary text and is what gets read.
Layout choices that matter
Single column is the safest shape, and Resume Clean is the one to pick if you care more about parsing than about the look. It's also the only layout with the one-click English download.
If you prefer a two-column layout on screen, keep it and export through print — the file you get is already straightened.
Keep the section names conventional
Section titles are editable here, which is a good feature and an easy way to hurt yourself. "Experience", "Education", "Skills", "Projects" are the words parsers are built around. "Where I've Been" is charming and unparseable.
The same goes for dates: a start and end month per entry, in the fields provided, so they come out formatted consistently rather than buried in a sentence.
What nobody can promise you
There are many tracking systems, they behave differently, and none of them publishes its parser. Anyone selling an "ATS score" is scoring your resume against their own guess.
What you can do is concrete: a text-based file, one clear reading order, conventional headings, real dates, and the same wording the job posting uses for the skills you actually have.