Is IE Usage Share Collapsing?

je ne sais quoi writes "Net Applications normally releases its statistics for browser and operating system usage share on the first of every month. This month, however, the data has not shown up — only a cryptic message stating they are reviewing the data for inexplicable statistical variations and that it will be available soon. Larry Dignan at ZDNet has a blog post that might explain what is happening: Statcounter has released some data that shows a precipitous drop in IE browser use in North America, to the benefit of Firefox, Safari, and Chrome. At the end of May, StatCounter shows IE usage share (for versions 6, 7, and 8 combined) at around 64%; at the beginning of June it is now about 56% — an astounding 8% drop in one month. We should keep in mind the difficulties in estimating browser usage share: this could very well be a change in how browsers report themselves, or some other statistical anomaly. So it will probably be healthy to remain skeptical until trend this is confirmed by other organizations. Have any of you seen drops in IE usage share for Web-sites you administer?"

Comments

typo in summary

[je ne sais quoi]

Hi there, submitter here. I left a typographical error in the summary. "in the beginning of June" should read "in the beginning of July". Oops, sorry about that.

Re:typo in summary

[boaworm]

I always thought that was the name of the guy who built Data. Wasn't it? Dr. Noone Young Sung?

Doctor Noonien Soong

Re:typo in summary

[PPH]

Listen, sonny. In my day we didn't have tags. We had to stand on street corners, ranting about any topic that came to mind.

And stay the hell off my lawn!

Re:typo in summary

[capnkr]

As a "moron" who reads summaries *and* even TFA's, and who even looks further into some things (gasp), I noticed that NetApp's Market Share page for Operating Systems has the same lack of data so far for this month as does the browser page.

Maybe this is just a case of "Nothing to see here, move along..." until we find out they had some mundane reason they were tardy this month.

Re:typo in summary

[mcgrew]

<pedantry>
It's not "noone" or "no-one", it's "no one". Two words. Got it?
</pedantry>

Now get off my lawn, you dumb kids. And no, you can't have your balls back.

My statistics

[GoNINzo]

I run a somewhat largish non-technology site, and I saw yesterday:
40.91% MSIE 7.0
27.11% MSIE 6.0
14.60% Mozilla/5.0
12.98% MSIE 8.0

Everything else below .1%. So that's 81% MSIE, 14.6% Mozilla, and everything else in the remaining 4.4%.

Re:My statistics

[GoNINzo]

No not really, we don't really track over time, though I have that capability. I do know that the MSIE 8 usage is way up, and really hitting the MSIE 7 usage more than anything else. The Firefox is about where it usually is.

Re:My statistics

[asa]

2001 called and said you can't use that tired old argument anymore. The default install of Firefox since 2.x (I believe) does not spoof IE in the user agent string. Firefox being the largest market share aside from any version of IE, the weight given to any other browser would be a statistical blip at best. In fact, if I remember correctly Konqueror in KDE3 and 4 actually spoofs Gecko by default. And Opera stopped spoofing MSIE after 6.x, IIRC.

No version of Firefox ever spoofed as IE. Safari has a "like Gecko" in their user agent string, and Opera spoofs only for specific sites.

Re:My statistics

[IntlHarvester]

As others said, forget spoofing.

However, ad blockers break the data collection for most analytics system. So it is likely that Firefox is being underreported, just because the of the popularity of ABP, NoScript, various cookie blockers, and so on.

Re:My statistics

[annodomini]

Hmm. You do realize that Safari reports itself as Mozilla/5.0, right?

Here's mine:

Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10_5_7; en-us) AppleWebKit/530.18 (KHTML, like Gecko) Version/4.0.1 Safari/530.18

They do this because various websites sniff for various browsers, and they want to show up as much like Mozilla/Gecko as possible. If your user agent parser isn't very smart, it might miss the Safari/530.18 part of that user agent string.

Of course, another possible explanation is that you work for a dental insurance company, for whom the most common users of the website are likely dental receptionists (for submitting claims), followed by people in HR (for signing up for services and looking up services on behalf of employees), both of which groups likely use only Windows machines.

No drop off here

[Anonymous Coward]

We've seen no major drop off, just a steady and slow decline. We track over 15 million users a day across the sites we manage here in the UK (mainly council properties).

It's because IE 6 support was droped on some sites

[denis-The-menace]

There are a few sites where IE 6.0 displays things badly because the web master stopped kludging for it.

Slashdot.org
some parts of Google.
(Help me here!)

Joe-six paks noticed this and has found out that he has options...

Looking from multiple angles

[asa]

If you look at the longterm trends reported by Net Applcations, something that StatCounter doesn't offer, it's hard to conclude that anything dramatic has just happened.

http://weblogs.mozillazine.org/asa/archives/2009/06/historical_view.html

These longer trends are steady and smooth and there's nothing that's happened in the last couple of months that would cause IE to fall off the cliff.

That being said, there is a lot of churn in the various browser versions. IE is really a collection of browsers with measurable share, IE 6, IE 7, and IE 8. Looking at these versions, it's clear that a lot is happening.

http://weblogs.mozillazine.org/asa/archives/2009/07/a_browser_prediction.html

It's likely that IE 7 and IE 6 will fall to under 10% global share by the end of this year and that IE 8 will grow to approximately 40%. That would give IE 60% overall, Firefox about 25%, Safari about 10%, and "other" would hold the remaining 5%.

1% maybe...

[mwhahaha]

On the two sites I have access to for this info IE dropped about 1% for May vs June. One site (~19M visitors a month) it was 57.91% vs 56.64%. The other (~132M visitors) it was 60.17% vs 59.40%. I always question these sort of numbers because browser usage is very closely tied with demographics, and I wonder just what sites are they using to get them...

Re:It's the iPhOnE!

[asa]

iPhone/iPod browsing makes up about half of one percent of Web usage. Desktop Safari makes up about 10% of Web usage. Firefox makes up about 25% of Web usage. I don't think the iPhone is having quite the impact you think the iPhone is having.

Please let this be true!

[mcrbids]

Not only would this change be welcome, but it would force Microsoft to "play ball" with the standards for HTML rather than roll their own and mark all the bug reports "will not fix".

Take a look at the history:

1) Microsoft is all about selling stuff on CD-ROM with the marketing vision "Information at your fingertips".

2) The Internet happens, and overnight, Netscape is a raving success because it actually PUT information at your fingertips.

3) Billy boy issues a memo to the whole company to turn as fast as possible to support the Internetz.

4) IE comes out - first a sucktacular mess, and finally almost livable around IE 5 or so.

5) IE 6 comes out, Netscape crumbles.

6) Netscape goes underground at AOL who throws a few developers at it while using it to negotiate a link on the Desktop. IE Dominates so tremendously that it's the platform of choice simply because it's installed everywhere.

7) Microsoft stops doing anything for half a decade. (whistle whistle)

8) Navigator continuously improves, finally re-emerging as Phoenix/Firefox. Suddenly, Microsoft's browser looks like a 5-year-old pile of cruft that's difficult to program for.

Suddenly, Microsoft will give a shiat. They might finally fix the things that developers!developers!developers! have been whining, bitching, complaining, and screaming about all these years.

Irony: "Free Internet Exporer 8" ad at the top while I type this message!

Re:Proliferation of mobile browsers...

[badasscat]

My stats only count desktop browsers and I am at 52.4% for all versions of IE. And I don't run a tech-heavy site or anything, I run a site selling Japanese clothes. (http://www.tokyorebel.com)

Firefox 3.0 is at 35.6%, 3.5 is at a surprising 0.6%, but then it's new. (And thank God, because some of my CSS is totally messed up in 3.5.)

Actually now that I'm looking, I do have a stat that says "iPhone" which is at 0.2%.

Re:It was to be expected

[Swizec]

I love a bad economy, it forces people to be less stupid.

But apparently a large portion of your business was relying on the fact people are stupid. Now what?

Re:Proliferation of mobile browsers...

[asa]

All mobile browsing combined probably doesn't break 1% of Web usage. The chances that any large-scale swing has anything to do with mobile browsing are very, very, very slim.

~20% here, and still in decline

[Enleth]

That's for a major Polish website devoted to a popular, long-running game series. The userbase is indeed a little more tech-conscious than the average Internet user around here, but not by much - just a few power gamers and techies, lots of "casuals". Nevertheless, IE was at ~70% in 2004, ~50% in 2005 and so on down to ~25% in the late 2008 and ~20% now. Right now it's kind of stabilizing (but still falling) and I don't forsee it falling below 15% anytime soon, but I'm starting to suspect that by the end of the year, Opera might overtake it (16% and rising, mostly ex-Firefox users right now).

We're not actively doing anything anti-IE or pro-FF/Opera (well, maybe except that IE is getting all the CSS/JS bugfixes lats, but that's *because* it's so low in the stats - we can afford letting the IE support lag behind), so it's mostly an outside trend, I think.

All the statistics I'm basing this post on were generated by Google Analytics, by the way.

Re:~20% here, and still in decline

[shutdown -p now]

I have no good explanation, only some guesswork. Computers came to the Soviet bloc later, and weren't even luxury items, but were only available at universities and at some workplaces. Because of that, a certain hacker-like mentality arose around them - quite similar to what happened in the West a decade or two earlier, but this time it was an epoch of PC already, and therefore the knowledge and skills were centered around that. The result is that CIS and Eastern Europe have an abnormally high concentration of DOS/Windows power users, who aren't afraid to mess with their system - and who historically prefer browsers other than IE for obvious reasons. Opera in particular was so successful because it was there (and more stable and feature complete) before Mozilla, and because the fact that it was then a commercial, for-pay product was entirely irrelevant in a software culture where using licensed software is considered eccentric.

One of the Opera guys had a market analysis of their market share recently, with essentially the same results as what GP describes for CIS and Eastern Europe (there's more there if you care to follow the links deeper). And Here is a map that shows the distribution visually (it's an interactive SVG, and Firefox screws it up, so you'll need Opera or Chrome to view it properly - but for the most part it just shows that the more eastern the country, the higher the Opera usage, peaking at ~50%).

In general, you can assume that wherever Opera is popular, so is Firefox (though it's interesting that Opera actually overtakes Firefox in Russia).

w3schools doesn't show anything

[Xtifr]

W3 Schools which has an admitted alternate-browser bias does not show any sort of abrupt drop-off for IE, and if anywhere were going to, I would think it would be this site. In fact, it shows Firefox dropping for the first time since September of last year (when Chrome was initially released), but only half a percentage point. IE7 is losing ground to IE8 rather quickly, but IE6 actually gained a half a percentage point since May. Chrome is also up another half a point, and nothing else really had enough movement to be worth mentioning (Safari up a tenth, Opera down a tenth).

My Stats Disagree

[Bill Dimm]

The stats for MagPortal.com (should be fairly unbiased) are not showing a drop in MSIE of that magnitude. Here is a comparison going from the last week in May to the first week in July:

MSIE: 66.10% -> 64.34%
Firefox: 25.71% -> 27.41%
Safari: 5.90% -> 5.61%
Chrome: 2.29% -> 2.65%

Re:Proliferation of mobile browsers...

[PaintyThePirate]

You think that's something? I host a SpongeBob fansite, and even it has 40.38% for Firefox and 47.90% for IE from June 6th to July 6th.

Looking at the data for the same period in previous years, I'm seeing:
2008: 63.26% IE and 31.49% Firefox
2007: 72.85% IE and 23.22% Firefox
2006: 77.60% IE and 17.77% Firefox

That's with 20,000+ visits in each period, so it's more than just noise.

Re:hmm.... it's summer?

[R3d M3rcury]

European nations require companies to give employees more paid vacations--4-6 weeks on average. Some companies pretty much shut down during the summer months. In the US, you tend to get your two weeks and that's about it.

Re:Proliferation of mobile browsers...

[Jason Levine]

You could use IETab for the sites that still need Internet Explorer. It can be set up so that the tab automatically uses IE for certain websites. The other sites will use FireFox as normal and users won't need to worry about firing up a second web browser. Then, if you update a web application so that it doesn't require IE6, you can remove that site from IETab's list. Users won't need to change their habits at all, but will get the FireFox rendering engine.

Re:It's because IE 6 support was droped on some si

[Atario]

It is driving me up the wall that an article page looks different if you go to it from the front page vs. hitting the headline embedded in the display controls. One goes to http://foo.slashdot.org/article.pl?sid=yy/mm/dd/idnumber and the other goes to http://foo.slashdot.org/story/yy/mm/dd/idnumber/Hyphenated-Article-Title . And the latter sucks.

Re:a fool to run Windows XP on a daily basis

[smash]

• Did you take their admin away?
• Did you enforce group (or local) policy on their machine?
• Did you apply any of the pre-built microsoft admin templates for security?
• Did you make use of IE's security zones

• No? Then you haven't tried all ways of trying to secure the box. XP (or any windows really) can be configured to be pretty sure, you just need to know shit from chocolate, pull your finger out and actually DO IT.

  Do I like doing that? No. Do I do that? No. But if i was desperate to secure a windows box from idiots, this is the path i'd be going down...

Re:Proliferation of mobile browsers...

[Kalriath]

Yes, actually, there is. Microsoft's BUILT IN InPrivate filtering by default manages to block every ad and tracking site (it works by auto-blocking things that are common to lots of sites) and you can set it to manual mode to block only things you don't want.

Also, IE7Pro (third party) goes the extra mile and is pretty much AdBlock, FlashBlock, and NoScript in one.

Re:a fool to run Windows XP on a daily basis

[IntlHarvester]

Disabled NoScript so the web would "work".

That sounds completely reasonable, disabling scripting does in fact make sites "not work".

Why are you foisting an extension for hardcore goatporn browsers onto regular corporate users?

Re:Now...

[DECS]

Yeah, Microsoft promises not to sue users to propagate Silverlight creep on Linux, as long as they are content to use an old version.

Silverlight = MS Flash: replacing the open web with a closed binary that only works well on Windows.

Ogg Theora, H.264 and the HTML 5 Browser Squabble

Net Applications, eh?

[notrandomly]

Not exactly the most reliable statistics. How can a browser have a higher market share than another browser with 3x as many users? Yeah, Net Applications reported that.