It’s time for the browser stats for October; as always according to StatCounter. I’m going to follow the desktop market too, for a while, since big changes are brewing.
Your donation for keeping this series up and running would be much appreciated.
Something big has happened on mobile: it turns out StatCounter has severely under-counted UC’s market share. It has not 1% but 5% of the global market.
Whose market share? UC’s. It’s the Chinese proxy browser with its own rendering engine. So far everybody assumed it was big in China but pretty much non-existent elsewhere, but now it turns out to also be pretty popular in India, Pakistan, Saudi Arabia, and a few other countries.
The problem was that the Chinese UC announced itself as “UCWEB” while the international version used “UC Browser.” StatCounter counted the first one correctly, but didn’t know about the second string.
I found out a while back and notified StatCounter. They made the change in late September, and while I expected to see some changes in the Chinese browser market I was totally unprepared for the large share it had in other countries.
The big loser seems to be Nokia. If StatCounter does not recognise a browser it assumes it’s the default browser of the device, and since most cheap devices are Nokias, UC Browser was counted as Nokia.
No more. UC +4, Nokia -4. That’s quite a change.
This episode proves that mobile browser statistics need constant care and attention, and lots of testing. I encourage you to send your obscure devices to StatCounter’s test page and warn them if they’re wrong. They like this sort of feedback and will amend their browser detection script.
In other news Safari once again jumps ahead of Opera, while BlackBerry loses some share and Android wins a point. NetFront loses one point, too, and that’s unusual. Its 4% is rock-solid ever since I started to keep track of mobile browser stats. Have some UCs been miscounted as NetFronts? Or is it genuinely going down?
The mobile browser market has shifted to a three-horse race at the top, followed by several browsers between 3 and 12%, followed by the long tail. It’s the race at the top that will interest most people, but what’s happening down the list is also interesting. One-third of all mobile web users do not use a top-three browser, after all. (On desktop the corresponding figure is only 9%.)
|Browser||October 2011||ch||September 2011||Remarks|
|Safari||23%||+2||21%||iPhone and iPod Touch|
|Opera||22%||0||22%||Mini and Mobile combined|
|UC Browser||5%||+4||1%||Chinese proxy browser|
|Obigo||1%||0||1%||For LG phones as well as Brew MP. Version 10 is WebKit-based|
|Samsung||1%||0||1%||Samsung’s non-Android, non-Jasmine, non-Dolfin browsers|
|Jasmine||1%||0||1%||Samsung NetFront-based and early WebKit-based|
|WebKit||58%||0||58%||Safari, Nokia, Android, Dolfin, 10% of BlackBerry|
|Mobile||7%||0||7%||Mobile browsing as percentage of all browsing|
On desktop things are changing, too, though at a more glacial page than on mobile. Basically Firefox and especially IE are losing some market share to the three smaller ones. Chrome’s upswing is most clearly visible, but Safari and Opera are also gaining little bits of market share — though usually far less than a point per month, so that it’s not easily visible in the table.
It could be that next month Chrome will overtake Firefox as the second-largest browser, and if current trends continue it will even overtake IE in about a year or so.
|Browser||October 2011||ch||September 2011||ch||August 2011||ch||July 2011|
Incidentally, StatCounter adds tablets to the desktop browsers, and they account for about 1% of all hits. That’s still way less than mobile, but tablets are growing at a prodigious rate. Of course the iPad takes about 90% of all tablet hits, and I think it’s actually the growth of tablet browsing, rather than Mac browsing, that causes Safari’s upswing.
Let’s hope StatCounter will split off tablets from desktop in the near future.
The table gives the global shares of the four device classes. I’ll expand on this in a future post.
I’ll be around at the following conferences: