blufive: (Default)
[personal profile] blufive

It’s a while since I wittered about browser stats. Which is a bit of a shame, because they recently got really interesting again.

Background info: these are stats from websites targeted at the general population of the UK (mostly car insurance, for anyone who cares that much). These sites are not “brochureware” search engine landing pages, they’re full-on quote-and-buy systems that can only be accessed after clicking through the initial marketing blurb, so they’re (mostly) only going to see the real core of people who are seriously considering actually buying stuff, rather than casual surfers. I have to be a little circumspect because this isn't really published data, but I figure if I've anonymised it this much, I shouldn't cause any problems.

These numbers are visitor counts based on the last seven days' data across a few sites - the total is about 40-50K visits. I know that's not a big number for this sort of data, but remember: seriously interested potential customers (Average page hits per visit is about 4-5, and those are big, heavy, lots-of-mandatory-form-fields pages) rather than every passing Tom, Dick and Harry one-page-hit-wonder.

The numbers are ranges because I've not tried to normalise/average them at all - so "25-30%" means one site had about 25%, another site had about 30%, and the others were somewhere in between.

Firefox 10+10-15%
Mobile Safari7-10%
Desktop Safari3-5%
Older Firefox2-3%
Android 2.2+1-1.5%

So... IE6 is dead. IE7 is in a death spiral.

5 months ago, IE8 was comfortably top dog, IE9/Chrome were neck-and-neck behind Firefox, and IE7 was a close fifth place.

The main reason for the major switcheroo appears to be that, at some point in February this year, Microsoft took the gloves off and set windows update to “latest IE, for everybody, extreme prejudice, NOW!”. That caused IE9 to eat a huge lump of IE8’s share in a matter of weeks. The only holdouts on IE6/7 are probably corporate shops that are trying hard to block the update – mere user apathy is no longer a factor.

I think the only reason IE8 still has significant numbers at all is that it’s the best WinXP can run. So, in the near-ish future, IE8 is going to be the remaining “fossil” browser, used only by the rump of people still on WinXP who don’t use Chrome/Firefox. On the corporate front, that is ceasing to be an issue, as even the slow movers are now upgrading in earnest to Win7.

Chrome has also been on a minor rampage for the last 12 months, stealing user share from everyone, though that seems to have levelled off a bit now. I suspect this is mostly down to the Chrome download adverts that get rammed down my throat every time I even glance at a google-run website, and the number of downloaded apps that attempt to piggyback-install it these days. Grr.

Firefox has (mostly) gotten over the bumpy switch to the rapid release cycle; for much of the last 12 months, there’s been a long tail of users spread across versions 3.5 to ${LATEST}. That’s now settled down to about a 60-40 split between 12.x and 10.x (the new “long term support” version) with a much smaller group of stragglers on sundry other versions.

Mobile is now 10% of browsers, easily. We’ve got individual sites clocking 12-13%. In general, these appear to be split in a ratio of about 40/40/15/5 iPhone/iPad/Android/other. The iPad/iPhone are mostly iOS 5.x with a few on 4.x; the phone/pad split is about 50-50, possibly leaning towards the pad a little. Androids are about a 55/30/15 split between 2.3 (Gingerbread), 2.2 (Froyo), and newer versions (Honeycomb, ICS).

(I reckon these mobile numbers are affected a lot by the fact that these sites are based around huge wodges of data entry, which is pretty painful to do on a 100mm screen, however you slice it)

TL;DR version: everything has changed around in the last 6 months. The days of “one major browser and a couple of minor ones” are well and truly over - the lead pack is now IE9, Chrome, IE8 and Firefox, with Mobile Safari a little way behind. Mobiles are now a significant fraction of the general browser population.

Note: boilerplate advocacy of "Browser X" in comments will be stomped and/or mocked. I don't want to moderate a Holy War.

Date: 2012-05-17 00:39 (UTC)
From: [identity profile]
Amusingly I am wearing the IE6 flatline t-shirt Microsoft sent me recently...

Date: 2012-05-20 18:42 (UTC)
From: [identity profile]
Stuff like the "death to IE6" campaign is one of the reasons I'm much nicer about the IE team than I was 8-10 years back. I'm still not a fan of IE9, but I can't say it's a bad browser, nor can I say that the IE people aren't trying... The big update push this year has been another thing worth applauding, and the effect really was quite startling when it first showed up in our stats.

Date: 2012-05-20 18:02 (UTC)
From: [identity profile]
My TL;DR; I can just fixup the rendering on our product on IE8+ then?

Date: 2012-05-20 18:38 (UTC)
From: [identity profile]
It would depend on the stats of the site in question, and what the official line is (or how much you think you can deviate from same). IE7 is still at about 5% for us, and therefore still gets some effort (read: as little as we think we can get away with) from us.

Certainly you should be paying more attention to mobile safari than IE7.


blufive: (Default)

April 2017

234 5678

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated 2017-09-24 11:58
Powered by Dreamwidth Studios