What are the most-viewed fonts in the world?

Stephen Coles

If you agree that webfonts are the most viewed fonts, the best measure was just released: HTTP Archive’s 2024 Web Almanac.

Family	desktop	mobile
Roboto	15.2%	2.7%
Font Awesome	10.4%	12.4%
Noto Sans JP	6.1%	5.7%
Open Sans	5.6%	6.8%
Poppins	4.7%	5.8%
Montserrat	3.3%	3.9%
Lato	3.2%	3.8%
Noto Sans KR	1.6%	0.8%
Source Sans Pro	1.4%	1.7%
Noto Serif JP	1.2%	1.4%
Proxima Nova	1.2%	1.2%
Raleway	1.2%	1.4%
Inter	1.0%	1.1%
icomoon	0.9%	1.1%
Oswald	0.7%	0.8%
Ubuntu	0.6%	0.8%
eicons	0.6%	0.8%
Barlow	0.6%	0.7%
Rubik	0.6%	0.6%
NanumGothic	0.6%	0.3%

Mark Simonson

The fonts in this list are fonts served as webfont resources. Fonts like Arial, Helvetica, Verdana, Georgia, etc.—in other words, web-safe fonts, the fonts that are commonly included on computers and therefore don’t need to be served—are left out by definition. I would guess that web-safe fonts are viewed much more than any of the fonts in this list. Unless you assume they are not used much anymore because most sites have switched to using webfonts.

Thomas Phinney

Most websites have indeed switched to webfonts. According to the same above-mentioned Web Almanac (https://almanac.httparchive.org/en/2024/fonts), 87% of websites served to desktops use hosted webfonts (85% for mobile).

UPDATE: that includes self-hosted webfonts

Mark Simonson

I guess I missed that part.  

jeremy tribby

interesting methodology, it resulted in a list that is a little different than the GF analytics page. "most viewed" is tricky as a metric, as it doesn't weigh in on the humanity of these viewers; more and more bots these days scrape content using residential IP addresses and normal-looking user agents, it's tough to know sometimes
well, in any case, note to self: merge that bug-fix branch of barlow

bramstein

The 87% is overall use of web fonts, including self-hosted fonts. About 13% of the analysed pages used no web fonts at all.

We did take a look at fonts referenced in the CSS font-family property (i.e. web fonts and locally installed fonts), but I felt the data needed more cleanup than we had time for (primarily because using @font-face you can give a font any name, so the data contained quite a lot of similar-but-not-quite-the-same names and random values). There's also the question of fallbacks. For example, given a font stack of "Proxima Nova, Helvetica, sans-serif", the actual font used may depends on what is loaded as a web font and installed on the user's device.

With all that said, the data we collected for the CSS font-family usage is public, and maybe next year we'll figure out a way to present it.

https://docs.google.com/spreadsheets/d/1EkdvJ8e0B9Rr42evC2Ds5Ekwq6gF9oLBW0BA5cmSUT4/edit?gid=1422791765#gid=1422791765

jeremy tribby

ahh yeah that sounds hard to normalize the names. there are also forks of projects that have been renamed intentionally but without other modification, sometimes you'll see open source fonts that end up with names like CompanyName Sans, CompanyName Serif, etc