http://gigaom.com/mobile/carrier-data-confirms-it-half-of-us-now-owns-a-smartphone/
Globally, smartphone penetration grew to 38%
Lets take this opportunity to share some data?
My mobile % is 32.
High traffic site, 3 months average ยจ36% mobile.
Anybody have a regex to quickly grab mobile UA strings out of logs?
I'm betting that on average people under 25 spend the majority of their web time on their smart phones.
http://www.justindocanto.com/scripts/detect-a-user-on-a-mobile-browser-or-device
I have between 35% and 41% as mobile traffic on the sites I just checked. Oh and 0.1% on PS3.
What are your thoughts on mobile sites? I prefer the standard sites on iphone and ipad than dedicated mobile sites, they just irritate me.
I have a couple around the 15% and some others around 27%
The lower % are for hospitality and higher ones are for retail.
>What are your thoughts on mobile sites? I prefer the standard sites on iphone and ipad than dedicated mobile sites, they just irritate me.
I'm seeing more mobile sites now that I'm using Chrome on the ipad and they irritate me. Before Chrome, I avoided this problem with the iCab browser because it lets you spoof the user agent. What I **REALLY** hate, though is some damn pop-up blocking the page so that they can tell me that they have a free app for their site.
In a way, I don't think ipads/kindles/nooks should be lumped in with the mobile stats. They're closer to laptops than smartphones.
>In a way, I don't think ipads/kindles/nooks should be lumped in with the mobile stats. They're closer to laptops than smartphones.
Totally agree, and I wonder what the numbers would be to reflect that.
Thanks Jason, using those UAs I wrote a quick perl script to get an idea of what one of my sites is doing % wise. It looks like my total percentage of mobile traffic is way lower than what you all are reporting. I'm seeing about 2%. Its probably the site's demographic.
If anybody is interested here is the script:
#!/usr/bin/perl -w
$log = $ARGV[0];
open(LOG,"$log");
my @log =<LOG>;
close(LOG);
my $mobile = 0;
my $fixed = 0;
foreach $line(@log){
@split = split(/\"/,$line);
$ua = $split[5];
#print $ua . "\n";
if ($ua =~/bot|spider|wget|google|archiver/){next}
if ($ua =~/android|avantgo|blackberry|bolt|boost|cricket|docomo|fone|hiptop|mini|mobi|palm|phone|pie|tablet|up\.browser|up\.link|webos|wos/i) {
$mobile++
}
else { $fixed++ }
}
my $total = $mobile + $fixed;
my $percentage = $mobile / $total;
$percentage = sprintf '%.2f%%', 100 * $percentage . '%';
print "Total desktop impressions: $fixed\nTotal mobile impressions: $mobile\nPercentage of mobile impressions: $percentage\n"
It's very basic, but seems to be working. There are probably a few more bot-like UAs to skip.
Use the script via command line like this:
./script.pl /path/to/log
Quote from: Drastic on August 22, 2012, 07:41:40 PM
>In a way, I don't think ipads/kindles/nooks should be lumped in with the mobile stats. They're closer to laptops than smartphones.
Totally agree, and I wonder what the numbers would be to reflect that.
Agree on tablets, although now a third class of site is starting, tablet (or iPad) specific versions of sites which look a bit like flipboard.
I do prefer to get a touch screen version of a site on smartphones, but I like having a link down at the bottom to switch to the web version if I want.
our sites vary 5% - 35% mobile, but on a lot of sites the mobile usage is very different from the other traffic - obvious ones being leisure where they want directions/opening times/confirmation of events quickly and ecommerce where people want to price compare/find the store. On the basis of giving people what they want, a lot of companies have to have a mobile (or at least suitably responsive) site or you'd get a huge bounce rate when they came into a category page from search.
I exclude tablets from this figures since they're different again in terms of usage and how people like to move around. We don't have anyone with a tablet specific site at the moment but that is going to be the way things go I think - or at least making the gallery style parts of sites more prominent when a tablet is detected....
Be interesting to see the age demographics against the number of mobile visitors.
I would guess the majority are kids fuelling the boom as they are more than likely have a smart phone given to them by mummy and dadddy than a PC/laptop.
More and more collage students will mean tablets, probably iPads. The move to etextbooks will fuel the use of tablets.
Just been playing in GA and come across a way to see how many impressions and clicks have come from smartphones.
You will need to link GA to GWT for this to work,
In GA go to Traffic Sources > Search Engine Optimization > Queries and add the secondary dimension Google Property
Then add an advance filter: include / Google Property / mobile (smartphone)
and hey presto :)
Does that help someone? or did everyone already know about this?
Google's adwords KW tool also has mobile/desktop market share per keyword which would be interesting for comparing different niches & countries.