By Craig Buckler

Browser Trends August 2011: Firefox 5 Grows by 400%

By Craig Buckler

July 2011 was a relatively quiet — perhaps vendors are enjoying a summer holiday following several months of frantic activity in the browser market.

However, that hasn’t stopped fickle users switching applications. The headline news is that the number of Firefox 5 users has more than quadrupled during the past 31 days. But you need to look at the full StatCounter statistics to reveal the true picture…

Browser June July change relative
IE 9.0 6.18% 7.27% +1.09% +17.60%
IE 8.0 27.67% 26.30% -1.37% -5.00%
IE 7.0 6.00% 5.45% -0.55% -9.20%
IE 6.0 3.72% 3.42% -0.30% -8.10%
Firefox 5.0+ 2.81% 14.28% +11.47% +408.20%
Firefox 4.0 14.04% 3.38% -10.66% -75.90%
Firefox 3.5+ 10.44% 9.36% -1.08% -10.30%
Firefox 3.1- 1.05% 0.94% -0.11% -10.50%
Chrome 20.67% 22.17% +1.50% +7.30%
Safari 5.07% 5.15% +0.08% +1.60%
Opera 1.74% 1.66% -0.08% -4.60%
Others 0.61% 0.62% +0.01% +1.60%
IE (all) 43.57% 42.44% -1.13% -2.60%
Firefox (all) 28.34% 27.96% -0.38% -1.30%

The table shows market share estimates for desktop browsers. The ‘change’ column shows the absolute increase or decrease in market share. The ‘relative’ column indicates the proportional change, i.e. another 8.1% of IE6 users abandoned the browser last month. There are several caveats so I recommend you read How Browser Market Share is Calculated.

While Firefox 5.0 enjoyed massive growth, it was released at the end of June so July provided the first full month of usage statistics. At the same time, Firefox 4.0 collapsed to 3.38% — in effect, the versions swapped places in the table.

Together, Firefox 4.0 and 5.0 grew by 0.81%. However, version 3.x and previous editions dropped by 1.19% giving an overall net loss of 0.38% for Mozilla’s browsers.

IE9 had another good month as more users migrated to the newer browser and, presumably, Windows Vista or 7 — the only OS’s which can run it. IE8 usage continues to drop by more than 1% per month. IE6 and 7 usage fell again and, while that’s good news, the browsers are still used by more people than Safari, Opera, and the “Others” combined. Overall, IE made a net loss of 1.13% — perhaps those users want to increase their IQ?!

Amazingly, Chrome usage grew by a further 1.5% in July — one of the largest monthly increases it’s ever experienced. Google is rapidly catching Mozilla and Chrome 12 overtook Firefox 4/5 again to become the world’s second most-used browser. That said, the rapid update schedules confuse the statistics. Firefox 5 will retake the #2 spot when Chrome 13 is released and the cycle will restart again.

Safari had a modest increase. Opera had a modest fall. However, it’s been a long time since Opera made a market gain.

Mobile Browser Usage

According to StatCounter, desktop browsers account for 92.98% of web activity. Mobile usage grew by almost 0.5% last month to 7.02%. This could be a seasonal anomaly and, interestingly, there was little movement within the mobile browser market. Perhaps that’s because we’re a few months away from present-buying season…

  1. Opera Mini/Mobile — 22.07% (down 0.74%)
  2. Android — 18.17% (up 0.92%)
  3. Nokia browser — 17.10% (down 0.56%)
  4. iPhone — 15.10% (down 0.12%)
  5. Blackberry — 12.30% (up 0.32%)

Android has overtaken Nokia’s browser to claim the #2 spot.

The iPhone continued to slide a little although it should be noted that the iPod Touch is at #6 with 4.85%. Apple enjoys a disproportionate quantity of technical news and development tutorial coverage. Geeks may love the devices, but be aware they account for less than 20% of the mobile browser market.

  • i cannot belive this. Chrome 13 is awesome. Its web dev tools has all the best features of Firebug and more

  • Been meaning to thank you for these monthly reminders to check the stats, Craig. As you say, the frequent version changes can confuse some comparisons, but that’s when StatCounter’s ‘browser’ graph becomes more useful than the ‘browser versions’. On that broader, more stable graph, it’s nice to see that in the UK at least, Chrome (all versions) has now toppled Firefox (all versions) as the second most popular browser. :-)

    • Thanks Pete. I find the figures interesting so I hope others do.

      I hadn’t noticed the UK-only figures: Chrome is now ahead of Firefox by around 0.5%. Google is marketing it heavily with billboards, newspaper and prime-time TV adverts.

      The European statistics are interesting too. IE and Firefox are evenly matched on 35% each, Chrome is at 21%, Safari 5% and Opera 4%. Safari has it’s strongest user base in northern America with 12%.

  • Anonymous

    The headline “Firefox 5 grows by 400%” may be more exciting than “80% of Firefox users upgrade to v5” but it is also misleading, in two ways. First, Firefox’s market share has actually fallen over the month, so while your headline isn’t factually incorrect in that v5 has shot up, the implication that Firefox has shot up is wrong. Second, it would be good to see numbers in there as well as market share. As one or two people on the forums are always keen to point out, 0.1% of market share might not sound much, but that’s still a lot of actual people around the world. A fall in market share could equally well at the same time be an increase in numbers … in other words, it might not be that anyone has turned away from Firefox, but just that more new surfers are using other browsers.

  • Craig, do yourself a favor and don’t write for a while. You are ruining your own credibility.

    The sum of market share for July is 27.96 while June has been 28.34. It means firefox market-share has been reduced.

    In IE Hox case, I would not blame BBC, CNN etc. because they do not claim to be web and internet professionals. You do!

    • Er, doesn’t the article point out Firefox’s market share has been reduced?
      If not, where did you find those figures?

      As for the IE hoax, when did I (or anyone else) blame another person or company? Many web professionals believed it was a real survey and articles appeared on other technical sites. I fell for it. Perhaps you didn’t — so where were your comments questioning it’s credibility?

      Sorry to disappoint, but you haven’t persuaded me to stop writing for SP. Of course, you’re free to stop reading my articles … but, given your comments, it appears you don’t read them anyway!

      • One reason I like DISQUS is that you can actually give people a chance to be honest about them self before they ask that from the others.

        That said, the title of the article and its content seems pretty clear and straight forward to me.

  • Chrome is not above all the other browsers in all and every aspect. Like any browser it has strong points and weak points.

    A weak point I discovered recently is the way it handles gradients. And it’s not a Webkit problem, since Safari manages them pretty well.

  • Anonymous

    It isn’t “too fast.” It is simply buggy. Saying that Chrome is “so fast it breaks connections” is just marketing nonsense. Don’t blindly believe Google’s marketing claims.

  • Why do you consider StatCounter to be terrible? That Wikipedia page uses them as a resource.

    The main reason I use their statistics is because they collate data from a diverse range of websites, publish their methods, and have the most detailed reports. NetApplications collate data from 40,000 sites. W3Counter use 49,000 sites. StatCounter use more than 3 million.

    Web statistics are not 100% accurate. Different collation methods yield different results.

    • Anonymous

      StatCounter is a terrible source because it has a lot of flaws. It doesn’t weigh different countries properly, so users in North America and Western Europe are given greater weight than users in countries where browsers like Opera are strong.

      Using Wikipedia is better because it at least combines several different sources, making it less terrible.

      • What makes you think that? Understandably, StatCounter may not always determine the correct country for a user’s IP, but a single visit is a single visit: regardless of the website, browser, or user’s location.

        The US and Western Europe only have more of an impact on global figures because they have a greater share of internet users. But that’s to be expected. You wouldn’t take Opera’s US figures (0.5%) and Norway’s figures (5%) then calculate an average market share of 2.75%!

        • Anonymous

          No, StatCounter simply does not weigh country data, meaning that one user in North America will give that browser a bigger increase in the stats than one user in Russia. They have admitted to that, and several people have pointed it out here and there. They even discussed it on Twitter.

          It is not about taking figures from one country and applying them to another. It is about how you look at the data from different countries. NA and WE are overrepresented, which means that their global stats are misleading.

          But never mind that. The bottom line is any single browser stats source is extremely unreliable. Wikipedia’s summary is still unreliable, but less so than StatCounter alone.

          • Why would StatCounter do that? It doesn’t make sense unless it’s because individuals in the US use the web more often than those in Russia. In that case, a US web user would perform more visits – but that’s fine, because market share measures usage.

            Please could you post the URLs of the sources you’ve seen.

          • Anonymous

            I don’t know why SC would do that. Why don’t you ask them? The bottom line is that their stats are flawed, which anyone with a clue knows.

            And yet the press and normal people keep blindly referring to it. It’s a crying shame.

            “Extremely poor stats are better than no stats” seems to be the excuse. Quite pathetic.

          • OK, but I really need to see some source materials and URLs detailing your concerns before I can consider another statistics provider.

    • Statistically there’s little difference in the accuracy of a pool of 3 million compared to that of 49 thousand, provided they are representative.

  • I guess it can be argued if that’s a good title or not. I thought it was a bit sensational, and up for interpretation.

    Nonetheless, thanks for reporting on this, and the analysis Craig. It seems like Chrome is steadily growing each month, is that a trend you’re seeing?

  • I guess it can be argued if that’s a good title or not. I thought it was a bit sensational, and up for interpretation.

    Nonetheless, thanks for reporting on this, and the analysis Craig. It seems like Chrome is steadily growing each month, is that a trend you’re seeing?

  • I guess it can be argued if that’s a good title or not. I thought it was a bit sensational, and up for interpretation.

    Nonetheless, thanks for reporting on this, and the analysis Craig. It seems like Chrome is steadily growing each month, is that a trend you’re seeing?

  • Anonymous

    These statistics are very interesting, as always. But I don’t think they are helped by sensationalist and misleading headlines like “Firefox 5 Grows by 400%”. The truth of the matter is simply that FF users upgrade versions very promptly, probably partly due to their nature and partly due to a very well-designed upgrade process built into FF.

    A lot of the commentary above seems to major on comparisons between specific browser/version combinations, which really obscures rather than illuminates the important points.

    Could I suggest you focus the presentation and analysis separately on two virtually independent aspects:
    1. Version changes within the same browser type, which illustrates upgrade patterns
    2. Changes in overall usage of browser types, regardless of versions, which indicates users making a choice to switch (or maybe being forced to switch by moving to a Mac!)

  • If it’s so “well known”, please provide some sources. Where are SC’s failings being discussed?

    PPKs article relates to mobile phones and it doesn’t mention anything about your accusation western bias?

    • Anonymous

      Look again. PPK’s article shows that there is something seriously wrong with their stats. They do not match up with reality.

Get the latest in Front-end, once a week, for free.