Browser Stats

I’m at An Event Apart in Chicago and Eric Meyer just said that browser statistics were “worse than useless.” More specifically, the only browser share numbers that matter are the one for sites you run, not what the web at large uses. Here’s our browser breakdown from 115 million visits to WordPress.com:

  1. 62.46% – Internet Explorer, sub-breakdown by popular request
    1. 64.10% – Version 6.0
    2. 35.17% – Version 7.0
    3. 0.28% – Version 5.5
  2. 30.74% – Firefox
  3. 3.83% – Safari
  4. 1.78% – Opera
  5. 0.52% – Mozilla

Just for fun, the operating system breakdown:

  1. 90.36% – Windows
  2. 6.73% – Macintosh
  3. 2.19% – Linux
  4. 0.03% – PlayStation Portable

66 thoughts on “Browser Stats

  1. Ewwww that’s too high for IE, I started converting my IE users to Firefox. It’s been a great revenue stream and helped the effectiveness of my coding.

    I feel that browser stats are actually quite useful. It provides a basis for which I actually design my sites. If I know my target audience, that provides me a platform for which I can optimize my site for that particular audience.

    I respect Eric and what his background brings to the table, but I disagree with him on that point.

  2. I wonder if those stats are close to overall market share or if wordpress visitors are more IE adverse than the general population – if it’s the former, then that’s a pretty strong/impressive showing for Firefox

  3. Actually the stats for WordPress.ORG are reversed:

    1. 52.73% – Firefox
    2. 36.77% – Internet Explorer
    3. 5.65% – Safari
    4. 2.89% – Opera
  4. I’m not surprised at the reversal for WordPress.org. Go take a look at Aaron Wall’s (SEOBook.com) stats while he has them on display. He’s a “techie” writer, and most of his users are using FireFox.

    And, I would say WordPress.com’s browser stats will probably be among the most accurate (unless you could take a look at Google’s stats)…

  5. The last time I was able to check meyerweb.com statistics (I think it was last year), I had more Firefox/Mozilla users than all versions of IE combined, much like the wordpress.org numbers.

    For those interested in more information on why I say “generic” browser statistics are “worse than useless”, see Don’t Care About Market Share (lawdy, that was written almost three years ago).

  6. That’s really interesting, especially when (I forget which speaker said it) it was said that WordPress is for geeks, whereas blogger is for everyone else. That idea works for .org but not for .com. Thanks for posting these.

  7. Google Analytics versions breakdown:
    For OS version go: Browser Capabilities > OS and click on the OS for a breakdown in versions.
    For Browser Versions go: Browser Capabilities > Browsers and click the browser name for versions.

    for http://wordpress.designpraxis.at I have for the top 5 OS:
    1. XP 83.05%
    2. Vista 10.81%
    3. 2000 4.56%
    4. Server 2003 0.95%
    5. 98 0.52%

    and for the top 5 browsers:
    1. Firefox 70.17%
    2. Internet Explorer 16.52%
    3. Safari 7.77%
    4. Opera 2.57%
    5. Mozilla 1.40%

    top 5 Firefox versions are:
    1. 2.0.0.6 88.35%
    2. 2.0.0.5 2.74%
    3. 2.0.0.4 2.51%
    4. 2.0.0.3 1.72%
    5. 1.5.0.12 1.59%

  8. Where’s the Konqueror browser? Are there really so few users with this neat KDE-browser, or did you merge it with Safari in the stats? (Konqueror is powered by the KHTML rendering engine which is also used by Webkit/Safari.)

  9. The wordpress.com statistics probably reflect more the general population than the wordpress.org ones since regular people aren’t usually tech savvy enough to install wordpress into a server, it’s just easier and just as effective for them to use a free service like wordpress.com

  10. While I agree that general browser statistics don’t mean as much as personal web stats for an established site, that doesn’t mean they’re worthless. You have to look at general web browser stats when you’re starting your web site design, because if your site doesn’t exist, you have no personal stats. You can look at general stats to figure out how important certain things are for your site design.

  11. Are there really so few users with this neat KDE-browser…?

    In my experience: yes. (But keep Eric Meyer’s article in mind.)

    My site, hyperborea.org, is seeing this for August through today:

    62.8% IE (26.4% IE7, 35.9% IE6)
    27.4% Firefox
    4.5% Safari
    1.8% Mozilla (prob. mostly Seamonkey)
    1.3% Opera
    0.9% Netscape (abt. 1/3 of these are “Netscape 4.0” — probably bots.)

    This is using AWStats, which lists totals and per-version breakdowns for IE, Firefox, and Netscape. Other browsers, like Opera and Safari, only list totals. It can actually recognize Konqueror, Camino, and several other browsers, but none of them manages to pull even 0.1% here.

    Still, Konqueror’s 1487 hits this month are better than OmniWeb’s 32, or iCab’s 19.

  12. “I feel that browser stats are actually quite useful. It provides a basis for which I actually design my sites.”

    I disagree. Standards provide a basis for design. Unless you are trying to identify an unusual user group (mostly mobile users, for example) any kind of generic statistics should be irrelevant.

  13. I’ve never really done anything with my browser stats. I just find it interesting to see who visits my site using what hardware and software.

  14. Strange to see that most of IE users are using IE6,and almost 90% Firefox users use 2.0.
    In my opinion browser stats are very usefull.
    Thanks for the info

  15. Not really surprising. Enjoyable though it is, blogging is still a partly ‘geeky’ pursuit. Geeks are more likely to use a non-IE browser, therefore geek-oriented sites like wp.com are likely to have a lower proportion of IE use.

  16. I am surprised to see IE scoring so heavily over Firefox on wordpress.com.

    Over last one month on my blog:
    Firefox: 69%
    IE: 20.45%
    Opera: 3.6%
    Safari: 2.0%
    Konqueror: 1.4%
    Others: Camino, Netscape, Galeon

  17. Standards are of course a basis for design. Unfortunately, some (ie) browsers don’t follow standards all the time. The browser stats just provide a reminder to not design your site for your firefox-using friends and also make sure the site looks good in IE 6 and 7 as well.

  18. im wondering if its possible to know what the screen resolutions are overall. that would be interesting i hear people are still on 800×600 and i shudder at the thought. but interesting enough im wondering if more people are using 1024×768 or 1280 x 1024….and what that breakdown is.

    nice work matt

  19. Once people start using IE7, Firefox will probably start falling even further. It’s hard to admit, but IE7 is actually a very good product (still not perfect, but makes Firefox less of a must-have).

  20. If we had to start “supporting Firefox” (i.e., using Web standards) when it hit 10%, does 36% IE6 usage mean we can stop “supporting IE6″ (i.e., defying Web standards)?

    I have about 16% IE6 on my sites and never lift a finger anymore to help them.

  21. Given the way IE handles CSS styles, standards come second to the statistics. Sure, I could design my site entirely in CSS styles that look beautiful in non-IE browsers, but why in the hell would I do that if 90% of my visitors are IE users unless I want to alienate them or force them into something they may not be comfortable with. In that case, I would be just as bad, if not worse, than Microsoft for forcing visitors to their site and resources to use IE.

    Knowing the statistics allows you to say, OK, I can use these standards and styles but not these, because it will alienate my users.

    You don’t have to like a browser to recognize a need to still code for it responsibly rather than annoy your visitors.

  22. For most blog, including WP.com and .org is representation of greek. So this will lead to the high percentage of Firefox users. If tracking on non-tech releating sites, just something els, IE will get higher score. The stat is directly relate to site content. I compared two of my sites, tech and non-tech, the stat is totally difference.

Comments are closed.