Is a Canva resume really ATS friendly?

We ran three layouts of the same resume through our own ATS text extractor and published the raw output: here is what software really reads in a Canva resume.

You spent two hours polishing a resume in Canva. It looks balanced, it looks professional, and you still hear nothing back. The question comes up on every job-hunting forum: does Canva break your resume for applicant tracking systems?

The short answer: Canva is not the problem, the layout is. And rather than just telling you that, we ran three layouts through the text extractor that powers our own ATS scanner, and published the raw output word for word.

The real question is not "Canva", it is "how many columns"

An ATS never looks at your resume. It flattens it into text, then hunts for sections, dates and keywords. It has no idea which tool produced the file: Canva, Word, LaTeX and our own editor all emit PDFs.

What changes everything is how the text is placed inside the file. A PDF contains no "left column" and no "right column" — it contains a list of instructions saying "draw this piece of text at this position". The order of those instructions is the order in which a machine will read you, and nothing obliges that order to match the path your eye takes.

That is exactly where two-column templates, which dominate Canva's resume gallery, come apart.

Our test: one resume, three layouts

How we ran it

We took a fictional resume — Claire Bennett, digital project manager, six years of experience — and produced three PDFs with exactly the same content. Only the layout differs:

1. Two columns, interleaved: sidebar on the left, main content on the right, with the text laid down one visual line at a time across the page.
2. Two columns, as blocks: same visual result, but the sidebar is one complete text frame placed before the main frame.
3. One column: everything in a single flow, standard section headings, contact details on one line at the top.

All three files then went through the text extractor HiredFolio uses to read imported resumes and to run ATS scans. The blocks below are its real, unedited output. Our extractor normalises whitespace, so what you are reading is the word order — which is precisely what an ATS has to work with.

Test 1 — two columns interleaved: the text turns to soup

CONTACT Claire Bennett claire.bennett@email.com Digital Project Manager +44 7700 900123 Manchester, UK PROFILE linkedin.com/in/claire-bennett Digital project manager with 6 years' experience leading website rebuilds and internal tooling programmes. SKILLS Project management WORK EXPERIENCE Agile / Scrum Digital Project Manager - Novaweb, Manchester Jira, Confluence 2022 - 2026 SQL, Power BI Led 12 rebuild projects with a combined budget of 1.4M GBP Coordinated 8 developers and 2 designers LANGUAGES Cut average delivery time by 30% English - native Spanish - fluent Web Project Coordinator - Atelier Blue, Leeds 2020 - 2022 Ran 25 client sites from wireframe to acceptance testing Introduced weekly budget tracking across the studio EDUCATION MSc Digital Project Management University of Manchester - 2020 BA Media and Communication University of Manchester - 2018

Look at what happens to the work experience. Job title, employer and dates no longer follow each other:

SKILLS Project management WORK EXPERIENCE Agile / Scrum Digital Project Manager - Novaweb, Manchester Jira, Confluence 2022 - 2026 SQL, Power BI

Between the section heading and the first role, the machine inserted "Project management" and "Agile / Scrum". Between the role and its dates, "Jira, Confluence". The fact that this job ran from 2022 to 2026 no longer exists anywhere in the file — it only ever existed in your eye.

The practical cost: an ATS that works out your tenure from the dates attached to each role finds nothing to attach. And a recruiter searching their database for "project manager" gets a record where the job title is buried in a list of tools.

Test 2 — two columns as blocks: content intact, order wrong

CONTACT claire.bennett@email.com +44 7700 900123 Manchester, UK linkedin.com/in/claire-bennett SKILLS Project management Agile / Scrum Jira, Confluence SQL, Power BI LANGUAGES English - native Spanish - fluent Claire Bennett Digital Project Manager PROFILE Digital project manager with 6 years' experience leading website rebuilds and internal tooling programmes. WORK EXPERIENCE Digital Project Manager - Novaweb, Manchester 2022 - 2026 Led 12 rebuild projects with a combined budget of 1.4M GBP Coordinated 8 developers and 2 designers Cut average delivery time by 30% Web Project Coordinator - Atelier Blue, Leeds 2020 - 2022 Ran 25 client sites from wireframe to acceptance testing Introduced weekly budget tracking across the studio EDUCATION MSc Digital Project Management University of Manchester - 2020 BA Media and Communication University of Manchester - 2018

This is much better: every section stays in one piece, and the dates sit right next to their role. But one detail is expensive.

The candidate's name, "Claire Bennett", only shows up at the 27th word, after the contact block, the skills and the languages. Parsers look for the candidate's identity at the very start of the document, because that is where it sits on the overwhelming majority of resumes. Here they meet the word "CONTACT" first, then an email address, then a list of tools.

That is the coloured-sidebar trap: visually, your name is at the top in large type. Inside the file, it is buried in the middle.

Test 3 — one column: read order matches intended order

Claire Bennett Digital Project Manager claire.bennett@email.com | +44 7700 900123 | Manchester, UK linkedin.com/in/claire-bennett PROFILE Digital project manager with 6 years' experience leading website rebuilds and internal tooling programmes. WORK EXPERIENCE Digital Project Manager - Novaweb, Manchester 2022 - 2026 Led 12 rebuild projects with a combined budget of 1.4M GBP Coordinated 8 developers and 2 designers Cut average delivery time by 30% Web Project Coordinator - Atelier Blue, Leeds 2020 - 2022 Ran 25 client sites from wireframe to acceptance testing Introduced weekly budget tracking across the studio EDUCATION MSc Digital Project Management University of Manchester - 2020 BA Media and Communication University of Manchester - 2018 SKILLS Project management Agile / Scrum Jira, Confluence SQL, Power BI LANGUAGES English - native Spanish - fluent

Name, title, contact details, profile, roles with their dates, education, skills, languages. The extracted text reads the way the resume reads. Nothing moved, and no date was separated from its role.

Same content, same design tool, three radically different results. The only variable is the number of columns.

One point matters here: in all three cases nothing was lost. Every word is present. So this is not a story about Canva deleting your text — it is a story about your words arriving in the wrong order. Which is more insidious, because nothing warns you.

What this actually costs you

Two beliefs get in the way of understanding the damage, one falsely alarming and one falsely reassuring.

"An ATS auto-rejects my resume." In practice an ATS ranks far more than it eliminates. A badly parsed resume does not trigger an automatic rejection: it sinks in the pile, it stops surfacing in the recruiter's keyword searches, and its experience is undervalued. The outcome looks a lot like a rejection, but the mechanism is different — and that is good news, because it is fixable.

"Recruiters like the design, so it evens out." On a posting with hundreds of applicants, the recruiter does not read the whole pile — they filter first. If you are not in the filter, your design is never seen.

It is also why our ATS scanner does not grade how your resume looks. When you drop a PDF into it, it works on the text that was actually extracted — the blocks above — and compares that text to the job you are targeting. An interleaved two-column resume scores lower because the content it hands to the analysis is objectively less usable, not because a template was blacklisted.

The ten-second test, on your own resume

You do not need a tool for the first diagnosis:

1. Open your PDF in any reader.
2. Select all (Ctrl + A, or Cmd + A), then copy.
3. Paste into a plain text editor or an empty email.

What appears is, near enough, what the software will read. Three possible verdicts:

• Nothing pastes: your resume is an image. This is the worst case — a PNG/JPEG export converted to PDF, or a fully flattened design. Not a single word is machine-readable.
• Text pastes but in the wrong order: you are in case 1 or case 2 above. Fix it.
• Text pastes in the order you wrote it: your layout is healthy, so focus on keywords instead.

Which Canva resumes survive

Canva exports real text when you download as PDF (standard or print). The problem is which template you picked, not the tool. Here is what separates a usable model from a booby-trapped one.

Two of these traps deserve detail.

Icons say nothing to a machine

A little envelope before your email, a handset before your number: to you the meaning is obvious. To the extractor, an icon is a drawing that produces no text at all. If your phone number is only identified by the icon in front of it, it becomes a floating string of digits. Write the labels out, or at minimum keep a recognisable number format.

Skill bars are just rectangles

A bar filled to 80% next to "Python" communicates no level whatsoever: the machine reads "Python" and nothing else. If the level matters, write it ("Python — advanced, 4 years"). You also gain with the recruiter, who reads a sentence better than a rectangle.

Fixing a Canva resume without starting over

You do not have to rebuild from scratch. In order of impact:

1. Switch to a single-column template. That is 80% of the gain. Canva's "simple", "minimalist" and "classic" models generally respect it.
2. Move your name and contact details to the very top, as text, before any other section.
3. Rename your section headings to the standard ones.
4. Put dates directly under the job title, not in a side column or on a timeline graphic.
5. Replace icons and rating bars that carry information with words.
6. Redo the copy-paste test. If the order is right, your layout is no longer costing you.
7. Mirror the keywords in the posting — once the structure is sound, that is where the rest of the score is decided. Our guide to resumes and ATS in 2026 covers that part.

A clean layout only carries what you put in it. If the bullets themselves need work — achievements, numbers, hierarchy — our guide to writing an effective resume picks up where the formatting fix stops.

A strategy that often wins: keep your Canva version for speculative applications, careers fairs, and anything you email or send over LinkedIn, where a human opens the file. Use a single-column version for anything that goes through an online application form.

What about Word, Photoshop or Figma templates?

The mechanism is identical, because it does not depend on the tool you started from. A Word resume with two columns built from an invisible table produces exactly the same scramble as our test 1. A Photoshop or Figma resume exported flat produces the worst case: an image, with zero readable words.

Remember the general rule rather than the list of tools: if the text does not copy out in the right order, your layout is costing you applications, whatever produced it.

Test your own resume

Everything above is a fictional resume. The only result that matters to you is your own file's.

HiredFolio's ATS scan is free and needs no account: drop in your resume, paste the job posting, and you get the match score plus what to fix — based on the text actually extracted from your file, not on how it looks. If the diagnosis confirms your layout is holding you back, HiredFolio can import the resume you already have, convert it into a clean structure and let you rebuild the formatting without facing a blank page.

Test your own resume for free

If you would rather cross-check a few opinions before reworking your layout, we reviewed the other tools on the market in our comparison of free ATS resume checkers.