Correction, 17 October 2026. Earlier versions of this article described a test of 12 fonts submitted to five ATS (Workday, Greenhouse, Lever, iCIMS and Taleo) and gave pass and fail scores for each font, such as "5/5 passed" and "2/5 failed". We have no records of that test, so we have removed it and the scores. We also removed claims about the servers and fonts those systems use, which had no source. The article now explains how text extraction treats fonts and gives the practical advice that follows from it.
"Use a clean modern font" is the most common, and least useful, font advice you'll get for resumes. The font name is rarely the problem. What matters is whether the file stores your words as real, readable text. Here is how that works, which fonts are the safe default, and the four font-related traps that can quietly turn your words into junk.
ATS-friendly fonts at a glance
| Font | Safe for ATS? | Good for |
|---|---|---|
| Calibri | Yes | Most resumes; the Word default, clean on screen |
| Arial / Helvetica | Yes | Plain and readable at small sizes |
| Times New Roman | Yes | Law, academia, traditional sectors |
| Georgia / Cambria / Garamond | Yes | Finance and senior roles that want a classic look |
| Google Fonts (Lato, Roboto, Open Sans) | Usually | Fine when exported to PDF from Google Docs |
| Decorative, script or icon fonts | Avoid | Headings only at most, never for contact details or body text |
The font name matters less than how the file stores the text. A resume in Calibri can still fail if it was exported as an image, and a resume in a less common font usually reads fine if the PDF holds real text. The quickest check is to scan the finished file and read what comes out.
How fonts affect ATS parsing
An Applicant Tracking System does not look at your resume the way a person does. It pulls the text out of the file and works with that text. So the question is not "does the system have my font installed?" but "does my file store each character in a way a text extractor can read back?"
- Word (.docx): the text is stored as plain Unicode characters. The font is just a name attached to each run of text. A parser reading a .docx gets your words whatever font you chose. A missing font changes how the document looks when someone opens it, not what is extracted.
- PDF: the file stores glyphs (shapes) plus, usually, a map from each glyph back to a character. When that map is present and correct, extraction is clean. When it is missing or wrong, the extractor has shapes with no letters behind them, and you get boxes, question marks or random symbols.
For a resume exported from Word, Google Docs or Pages in a standard font, the map is almost always correct. The trouble starts with the cases below.
The safe list (use these)
These are the safe default because they are installed almost everywhere, every mainstream editor exports them cleanly, and they read well on screen and in print. Among them, pick on taste.
1. Calibri
The default font in Microsoft Word for over a decade. If you can't decide, use Calibri.
2. Arial
Even more widespread than Calibri, with a slightly heavier look. For parsing purposes it behaves the same.
3. Helvetica
The original sans-serif workhorse. On Windows machines without it, it is usually shown as Arial, which only affects how it looks to the person opening the file.
4. Times New Roman
The classic serif. Some recruiters consider it dated, but it reads cleanly and parsers do not care how it looks.
5. Garamond
A slightly more elegant serif than Times New Roman. A popular choice for traditional MBA admissions resumes; see our MBA resume format guide for the school-by-school conventions.
6. Cambria
Microsoft's modern serif and a good substitute for Times New Roman. Available on most systems.
7. Verdana
Designed for screen readability and a little wider than Arial. Fine for parsing; it just takes more space per line.
The real risks (watch for these)
None of these are about a font being "modern" or "designer". Montserrat, Poppins, Lato or Open Sans can be perfectly safe when exported cleanly. The risk comes from how the file was made, and design tools and template sites hit these traps more often than Word does.
Ligatures
Many fonts can join letter pairs like "fi", "fl", "ff" and "ti" into a single glyph. Design tools often switch this on by default. Some text extractors read that joined glyph back correctly; others read it as one character or drop it, so "office" can come out as "of?ce" and "portfolio" as "portolio". The keyword is then gone from the extracted text. LaTeX users hit the same problem from a different direction, because pdfTeX joins "fi" and "fl" by default; the preamble fix is in are LaTeX resumes ATS-friendly.
Icon and symbol fonts
Phone, email and LinkedIn icons are often drawn with an icon font. To a text extractor each icon is just a character, frequently from a private-use area of Unicode with no meaning. It shows up as a box or a stray symbol next to your contact details. Word's default bullet symbol can do the same thing. Our scanner flags these characters because they are a common source of junk in extracted text; our Canva-style template test shows what icons turned into.
Wide letter-spacing
A PDF extractor works out where one word ends and the next begins from the gaps between letters. Headings set with wide tracking (a common template style for names and section titles) can come out as "S O F T W A R E" or split mid-word. If a heading or your name looks spaced out, set the spacing back to normal.
Fonts embedded without a character map
Some PDF tools, print drivers and "flatten" options embed fonts without the glyph-to-character map. The page looks perfect, but the extractor gets shapes with no text behind them. A resume saved as an image, or scanned, has the same problem in its purest form: no text at all. A quick test is to open your PDF, select all the text and paste it into a plain text editor. If what you paste is not what you see, a parser will not read it either.
Comic Sans, Papyrus, Brush Script
These usually extract fine. The problem is the reader: a recruiter will not take them seriously. Script fonts also tend to be hard to read at body size.
Special cases
Custom or licensed fonts
In a PDF, an embedded font with a proper character map extracts like any other. In a .docx, fonts are referenced by name, so the text is still extracted correctly, but anyone who opens the file without that font sees a substitute, which can shift your layout and page breaks. If you use a custom font, send a PDF.
Mac users, Helvetica Neue
The default Mac font. In a PDF it is embedded and reads fine. In a Word file opened on Windows it is usually shown as Arial, which can change line breaks. Plain Helvetica or Arial avoids the surprise.
Resume builders and design tools
Online builders and design tools (Canva and similar) often use web fonts, ligatures, icon fonts and letter-spaced headings together. Always run the copy-and-paste test above on the downloaded file, or scan it, before submitting.
Smart quotes
Word replaces straight quotes (') with curly ones (’) as you type. Curly quotes are ordinary Unicode characters and a well-built parser reads them without trouble; in our own parser benchmark, em-dashes and curly quotes came through cleanly. The small remaining risk is an older system that converts text to plain ASCII and turns them into question marks, giving "Don?t". If you want to remove that risk, switch them off.
Fix: in Word, go to File → Options → Proofing → AutoCorrect Options → AutoFormat as You Type → uncheck "Straight quotes with smart quotes." Then retype or find-and-replace the quotes in your resume and re-save.
Recommended font sizes
- Name (top of page): 16-20pt
- Section headers (PROFESSIONAL EXPERIENCE, EDUCATION, etc.): 12-14pt
- Job titles within sections: 11-12pt bold
- Body text and bullets: 10-11pt
- Below 10pt: parsers read it fine, but human readers may not. Avoid going below 9.5pt.
- Above 12pt: wastes space and looks unprofessional. Save large sizes for the name only.
Quick recommendations for different roles
- Tech / startup roles: Calibri or Arial (clean, modern, universally available)
- Finance / consulting / law: Garamond or Times New Roman (signals seriousness)
- Creative roles (when ATS isn't the gatekeeper): still use a safe font for the ATS submission, and send your portfolio separately
- Academic positions: Times New Roman or Garamond (the usual standard for CVs)
Don't optimize what doesn't matter
The choice between Calibri, Arial, Helvetica, Garamond and Times New Roman makes no practical difference to parsing. Pick whichever looks best to you and move on.
Don't spend hours debating fonts. Spend that time on bullet content, keywords and parsing structure (single column, standard headings, no text in headers or tables). Those matter far more than font choice.
What you should avoid is the combination that template sites and design tools often produce: ligatures on, icon fonts for contact details, letter-spaced headings, and a PDF export that may not carry a character map. That combination can look perfect on screen and still give a parser broken text.
Run a quick check: scan your resume free and see exactly what our parser extracted. If you see boxes (□), question marks where punctuation should be, words split by spaces or missing letters around "fi" and "fl", your font settings are the problem. Switch to one from the safe list, turn off ligatures and letter-spacing, and re-scan.
