Change of Font Size increased traffic by 100%

[dailypress]

Hello all,

Just wanted to share an interesting experience. I have a 5-7 page website which the domain itself and the content is based on a controversial subject.

The site is live since 2007 and has received a lot of strong backlinks. During the past few months it used to get around 50 visitors/day (over 90% of the traffic is from Google).

i.e.:
prev font size (main context): 9px
current size: 13px

With an increase in font size across the whole site (using CSS) the traffic doubled within 3-4 weeks.

The site in general looked more professional with smaller font however, the font size increase for the title, subtitle and context almost doubled my traffic.

I am just hoping this increase of traffic keeps on continuing...

[aakk9999]

This is strange unless there is some user metrics that changed as a result of your font change that resulted in Google sending more traffic.

Have you tried to analyse further? For example:
- have your ranking changed
- are there more keywords driving traffic
- is the increase of traffic all from Google
- do you have longer time on site, lower bounce rate and more pages per visitor
- has the change occured on or around the date where we know there have been some Google updates
- or is there a seasonal search influence or something that would result in visitors searching for the subject of your site more

It would be interesting to do a bit deeper analysis.

[lucy24]

prev font size (main context): 9px
current size: 13px

I should ### well hope so. NINE pixels?!?!?!

[Hoople]

If the visitor can't read your site they hit the back button. Perhaps it was your bounce rate dropping and more time on page as a result?

The visitors you had previously were likely gifted with eyesight best able to cope with the sub normal font size?

[aristotle]

I wouldn't have thought that this would make much difference.
Don't most people have their browsers settings set to override the code-specified font size and show their preferred size instead? In other words, the traffic increase you saw could have been due to something else and just by coincidence happened right after you made the font-size change.

[Robert Charlton]

It's a usability issue. 9px is ridiculously tiny, IMO. I would be interested in which user metrics you see changing.

This thread from 2006, when people had roughly the same visual capabilities they do now, might be of interest...

Using small sized text...
http://www.webmasterworld.com/forum5/7890.htm [webmasterworld.com]

tedster posted...

As a point of interest, 8px is the smallest font size that offers enough pixels to display mixed case text (which needs ascenders and descenders.) ...As a general rule, I've found that when you make the font size larger (say at least 12px) you can watch your stats go up.

The entire discussion is worth a read.

Incidentally, I've been noticing many sites recently turning to very large size fonts that look good at larger sizes. Font size does affect different fonts differently.

[phranque]

i've always heard the rule of thumb for font size was 1/3 the age of your intended audience.
so 9px is ok up to mid 20s which might rule out a large percentage of your visitors.

[lucy24]

Don't most people have their browsers settings set to override the code-specified font size and show their preferred size instead?

You've spent too much time among webmasters :) Most people don't even know that their browsers have prefs, let alone how to change them. Well, considering that a pretty significant chunk of people don't know what a browser is in the first place...

Paradoxically, a font size in explicit pixels is easier to override that something equally egregious like "50%" or "xxxx-small". Not all browsers can deal with computed sizes.

[Hoople]

I recently (based on feedback) changed a site I maintain for a non profit group to let the browser/user set the size of the base font (p) and scaled up the H3, H2, H1 from there.

The average age of the users is 70 (not a typo) and the 60-85 y/o's comprise 85% plus of the spenders on their events/fund raising trinkets.

I asked them face-to-face what sites they like the text size of and which they didn't. A majority of the sites they hated were the 'pixel perfect' sites made for 1280x1024 display, which all had the p set to 10px and below size.

[viral]

I am pretty sure Matt Cutts even mentions readability in a recent video. He doesn't talk about font size as I recall, but he does talk about readability and I assume font size would be a part of that.

As Hoople says above it may be an indirect consequence of bounce rate dropping.

You should go back through your analytics logs and see if you where ever hit by the Panda update. There could be a correlation between font size and panda. Something I have not considered before.

[mromero]

I noticed the same after changing templates.

The developer has for some time been saying that larger font sizes look better to him. I think he does not realize that his eyesight is getting shot like most middle-age people who spend hours on the computer, and is unconsciously designing with bigger size fonts ;o)

What some some designers do not get is that a good percentage of the population has a color vision deficiency (color blind). I think it is like 8 per cent of males worldwide.

The people who design search engine UI know this very well I think, reason why a good number of users have no idea what is an advert and what is a search result.

[lucy24]

i've always heard the rule of thumb for font size was 1/3 the age of your intended audience.

What some some designers do not get is that a good percentage of the population has a color vision deficiency (color blind). I think it is like 8 per cent of males worldwide.

Inevitable follow-up question: Is there a correlation between user age and preferred contrast level?

All I know for rock-solid sure is that my personal pages' background color is getting lighter and lighter. 15 years ago it was ABF; it's now CDF and climbing.

[dailypress]

It's a usability issue. 9px is ridiculously tiny, IMO.

Yes I agree. The site seemed congested but looked more professional with smaller font especially that my web designer aligned certain paragraphs with certain pictures. The font size increase ruined the layout but it seems much more user friendly and readable.

I need to give it a bit more time to figure it out. Will let you know soon...

[littlecubpanda]

I've got everything at around 13 or 14 size, mostly Verdana. My only concern is that it might look too big on an iPhone or other small mobile phone. Otherwise, I think bigger is better.

[dethfire]

To me 13/14 with verdana is really huge. WebmasterWorld looks like they use 10pt verdana and it reads fine.

[lucy24]

Sans-serif fonts "read" bigger than serif because they tend to have proportionally larger x-heights. Unfortunately the "font-size-adjust" property isn't very well supported yet :(

:: shuffling papers ::

http://www.w3.org/TR/1998/REC-CSS2-19980512/fonts.html#propdef-font-size-adjust

Here I've deliberately quoted the CSS 2.0 version, as it's less subject to change than the CSS 3.0 equivalent. (It isn't in 2.1.) Intentionally de-linked because the necessary fragment # breaks under auto-linking.

There's a Firefox add-on that identifies fonts used on a site. It only works with named fonts; if the code simply says "sans-serif" the add-on won't tell you what specific font your browser is using. Presumably if you know enough to add a plugin you also know enough to find your browser prefs, where you yourself can tell the browser what to use.

[edited by: lucy24 at 9:18 pm (utc) on Jul 2, 2013]

[Simsi]

I've got everything at around 13 or 14 size, mostly Verdana. My only concern is that it might look too big on an iPhone or other small mobile phone. Otherwise, I think bigger is better.

Shrink it a tad and use line-height to space it. That works well from a readability angle.

On the OP, my thought is that *if* there is a correlation between font size and ranking it is only that 9px was negatively impacting rank and 13ps is deemed acceptable.

[gouri]

I have body text that is coded size 2 (this is HTML; I am going to make it CSS later on, but it is size 2 HTML at the moment), but I think that it appears smaller than a size 2 on the web pages (when I copy and paste the text on the web pages onto Microsoft Word, it is a lot bigger in Word).

Could this be impacting me? Should I use size 3?

I have seen many pages on the Internet that appear to be using the equivalent of size 2 font.

[Dymero]

Depends on the font. WebmasterWorld is using 2, but it's using Verdana as the family, which is a larger font to begin with. Three, at least on Arial, appears to be equivalent to 13 or 14 pixels. I don't necessarily know that having it on 2 is affecting you, as it appears to be 11 or 12. But if your audience is having trouble reading the words and backing out of the site, I suppose it's possible.

On a sidenote, I didn't realize so many sites still use the font tag.

[viral]

Font tag is very old school. If I was involved with the Google algorithm I would consider that tag as a Panda flag. I am not saying good sites don't use it, for example WebmasterWorld uses it but it would have to eventually factor as a marker of a very old site that might be less relevant.

Anyway just saying..

[netmeg]

I don't use less than 16px on any of my personal sites. And a line height of 120% to 140%. And my sites do pretty well. If my pages weren't already too long, I'd probably take it up a notch, too.

It resizes just fine on smart phones and tablets, too. But my sites are responsive, so we can tweak it in the stylesheet.

[lucy24]

I don't think I ever specify an explicit font size. The whole point-- haha-- of CSS is that it cascades: 120% of this, 150% of that.

My default line height in paragraphs is 1.2. This is, among other things, enough that if a paragraph contains a sub- or superscript, the spacing won't visibly change in the affected line.

Font size 2 is the default so there's no earthly reason to say anything about it at all.

:: detour here to look at source code of present page ::

Oh, the horror. I assume it's all done in php and nobody has looked at the actual code since 1997. But then, I pepper my own php with explicit newlines so people who snoop can't tell which parts were hand-rolled.