(Image courtesy of Guy Moorhouse)
Earlier this month I dug into the analytics data to better understand what devices people are using when visiting GOV.UK. I thought I'd also quickly share headline data on what browsers, operating systems and screen resolutions we've seen over the past month across the whole site. I hope it's useful.
| Rank | Browser | Jan-14 | Jan-13 |
| 1 | Internet Explorer | 29.2% | 38.7% |
| 2 | Chrome | 27.5% | 21.8% |
| 3 | Safari | 25.0% | 20.4% |
| 4 | Firefox | 8.5% | 10.7% |
| 5 | Android Browser | 7.0% | 5.8% |
| 6 | Safari (in-app) | 0.8% | 1.4% |
| 7 | BlackBerry | 0.7% | 0.4% |
| 8 | Amazon Silk | 0.5% | 0.0% |
| 9 | Opera | 0.4% | 0.5% |
| 10 | Opera Mini | 0.2% | 0.2% |
We'll publish more detailed browser version breakdown data soon, though I can't resist sharing that Microsoft IE6 usage seems to have halved to 0.4% from 0.8% over the past 12 months. Note: this browser data combines both desktop and mobile versions.
| Rank | Operating System | Jan-14 | Jan-13 |
| 1 | Windows | 57.0% | 67.9% |
| 2 | iOS | 22.5% | 16.3% |
| 3 | Android | 12.0% | 6.6% |
| 4 | Macintosh | 5.8% | 5.9% |
| 5 | Linux | 1.1% | 1.1% |
| 6 | BlackBerry | 0.7% | 1.0% |
| 7 | Windows Phone | 0.7% | 0.3% |
| 8 | Chrome OS | 0.2% | 0.1% |
| 9 | (not set) | 0.1% | 0.5% |
| 10 | Series40 | 0.0% | 0.0% |
No real surprises here, with Android and iOS mobile operating systems eating into Windows desktop share.
| Rank | Screen Resolution | Jan-14 | Jan-13 |
| 1 | 1366x768 | 17.9% | 19.4% |
| 2 | 768x1024 | 10.4% | 7.7% |
| 3 | 1280x800 | 6.9% | 9.8% |
| 4 | 1024x768 | 6.6% | 9.8% |
| 5 | 1280x1024 | 6.6% | 7.0% |
| 6 | 320x480 | 6.3% | 7.3% |
| 7 | 320x568 | 6.0% | 1.0% |
| 8 | 1920x1080 | 5.2% | 4.2% |
| 9 | 1440x900 | 4.3% | 4.9% |
| 10 | 1600x900 | 2.6% | 2.5% |
Again, the story here is the rapid rise of "portrait" smartphone screen resolutions such as 768x1024 at the expense of traditional "landscape" desktop resolutions. Further evidence of the rapid shift to a wider variety of screen sizes as mobile device use takes off.
(Data sourced from Google Analytics for all visits to GOV.UK for the month up until January 20th. Sample size Jan 2014: 36.4m visits. Jan 2013: 28.4m visits.)
Update 5th Feb: A couple of you asked for more detail on devices, so here's the top 100 smartphones and tablets used on GOV.UK in January 2014.
| Rank | Device | % age visits |
| 1 | Apple iPhone | 11.01% |
| 2 | Apple iPad | 9.86% |
| 3 | Samsung GT-I9300 Galaxy S III | 1.33% |
| 4 | Samsung GT-I9505 Galaxy S IV | 1.15% |
| 5 | (not set) | 0.66% |
| 6 | Samsung GT-I9100 Galaxy S II | 0.53% |
| 7 | Samsung GT-I8190N Galaxy S III Mini | 0.44% |
| 8 | Google Nexus 7 | 0.34% |
| 9 | Samsung GT-P5110 Galaxy Tab 2 10.1 | 0.31% |
| 10 | Amazon KFTT Kindle Fire HD 7 | 0.29% |
| 11 | Samsung GT-I9195 Galaxy S4 Mini | 0.25% |
| 12 | Samsung GT-P3100 Galaxy Tab 2 7.0 | 0.25% |
| 13 | Samsung GT-S5830i Galaxy Ace | 0.24% |
| 14 | Samsung SM-T210 Galaxy Tab 3 7.0 | 0.23% |
| 15 | HTC M7 One | 0.21% |
| 16 | Samsung GT-N7100 Galaxy Note II | 0.21% |
| 17 | Apple iPod | 0.19% |
| 18 | SonyEricsson C6603 Xperia Z | 0.19% |
| 19 | Nokia Lumia 520 | 0.19% |
| 20 | BlackBerry 9320 BlackBerry Curve 9320 | 0.15% |
| 21 | Samsung GT-I9305 Galaxy S III LTE International | 0.15% |
| 22 | Samsung SM-N9005 Galaxy Note 3 | 0.14% |
| 23 | Samsung GT-I8160 Galaxy Ace 2 | 0.14% |
| 24 | Tesco Hudl HT7S3 | 0.13% |
| 25 | Samsung GT-N8010 Galaxy Note 10.1 | 0.13% |
| 26 | Samsung GT-P5210 Galaxy Tab 3 10.1 | 0.13% |
| 27 | Samsung GT-P3110 Galaxy Tab 2 7.0 | 0.12% |
| 28 | Sony C6903 Xperia Z1 | 0.12% |
| 29 | Sony C5303 Xperia SP | 0.11% |
| 30 | Sony LT30p Xperia T | 0.11% |
| 31 | Samsung GT-S5830 Galaxy Ace | 0.10% |
| 32 | Samsung GT-I8190 Galaxy S III Mini | 0.10% |
| 33 | Samsung SM-T310 Galaxy Tab 3 8.0 | 0.10% |
| 34 | Samsung GT-S5360 Galaxy Y | 0.10% |
| 35 | Samsung GT-N7000 Galaxy Note | 0.10% |
| 36 | Samsung GT-I9100P Galaxy S II NFC | 0.09% |
| 37 | Amazon KFOT Kindle Fire 7 | 0.09% |
| 38 | HTC PJ83100 One X | 0.09% |
| 39 | Google Nexus 4 | 0.08% |
| 40 | Nokia Lumia 800 | 0.08% |
| 41 | BlackBerry PlayBook | 0.08% |
| 42 | SonyEricsson LT26i Xperia Arc HD | 0.08% |
| 43 | BlackBerry 9900 Dakota | 0.07% |
| 44 | Samsung GT-N7105 Galaxy Note II | 0.07% |
| 45 | Amazon KFSOWI Kindle Fire HD 7 3rd Gen | 0.07% |
| 46 | RIM Z10 | 0.07% |
| 47 | SonyEricsson ST25i Xperia U | 0.07% |
| 48 | Samsung GT-S6810P Galaxy Fame | 0.06% |
| 49 | Sony ST26i Xperia J | 0.06% |
| 50 | Samsung GT-S5839i Galaxy Ace | 0.06% |
| 51 | LG Nexus 5 | 0.06% |
| 52 | Huawei Y300 Ascend Y 300 | 0.05% |
| 53 | Motorola XT1032 DVX | 0.05% |
| 54 | HTC One mini | 0.05% |
| 55 | HTC A510 Wildfire S | 0.05% |
| 56 | HTC 8S | 0.05% |
| 57 | HTC PJ401 One S | 0.05% |
| 58 | HTC 1000C Desire C | 0.05% |
| 59 | Sony C1505 Xperia E | 0.05% |
| 60 | RIM KBD | 0.04% |
| 61 | BlackBerry 9300 Curve 3G | 0.04% |
| 62 | Samsung GT-S5570 Galaxy Mini | 0.04% |
| 63 | Nokia Lumia 820 | 0.04% |
| 64 | Samsung GT-I9000 Galaxy S | 0.04% |
| 65 | BlackBerry 9360 Curve | 0.04% |
| 66 | Nokia Lumia 920 | 0.04% |
| 67 | Acer Iconia B1-710 | 0.04% |
| 68 | Sony ST23i Xperia miro | 0.03% |
| 69 | Nokia Lumia 620 | 0.03% |
| 70 | Samsung GT-I9500 Galaxy S IV | 0.03% |
| 71 | Motorola XT890 RAZR i | 0.03% |
| 72 | Sony C1905 Xperia M | 0.03% |
| 73 | Nokia Lumia 610 | 0.03% |
| 74 | Asus ME173X Memo Pad HD7 | 0.03% |
| 75 | HP Slate 7 | 0.03% |
| 76 | Google Nexus 10 | 0.03% |
| 77 | Nokia Lumia 625 | 0.03% |
| 78 | Samsung GT-S7500 Galaxy Ace Plus | 0.03% |
| 79 | SonyEricsson LT18i Xperia Arc | 0.03% |
| 80 | HTC Desire | 0.03% |
| 81 | Amazon KFTHWI Kindle Fire HDX 7 3rd Gen | 0.03% |
| 82 | Sony C2105 Xperia L | 0.03% |
| 83 | BlackBerry 9790 BlackBerry Bold 9790 | 0.03% |
| 84 | LG E400 Optimus L3 | 0.02% |
| 85 | Nokia Lumia 925 | 0.02% |
| 86 | HTC Desire HD | 0.02% |
| 87 | Sony ST21i Xperia Tipo | 0.02% |
| 88 | Nokia Lumia 710 | 0.02% |
| 89 | BlackBerry 9800 Torch | 0.02% |
| 90 | Acer A1-810 | 0.02% |
| 91 | Samsung GT-P5100 Galaxy Tab 2 10.1 | 0.02% |
| 92 | LG D802 G2 | 0.02% |
| 93 | HTC T328E Desire X | 0.02% |
| 94 | HTC Desire S | 0.02% |
| 95 | Samsung Galaxy Nexus | 0.02% |
| 96 | BlackBerry 8520 Curve | 0.02% |
| 97 | Samsung GT-P7510 Galaxy Tab 10.1 | 0.02% |
| 98 | Acer B1-A71 Iconia B1-A71 | 0.02% |
| 99 | Nokia Lumia 1020 | 0.02% |
| 100 | Amazon KFJWI Kindle Fire HD 8.9 | 0.02% |
(Data sourced from Google Analytics for all visits to GOV.UK for the month of January 2014. Sample size 47m.)
Follow Tom on Twitter and don't forget to sign up for email alerts
You may also be interested in:
The mobile question: responsive design
22 comments
Comment by Nao Yoshino posted on
Hi Tom,
Late to the party, but thanks for detail on devices (5th Feb). Is there any news on the break-down by browser? Quite keen to see the figures for IE7/8, and if possible, a break-down of browsers by OS across all platforms.
Thanks,
Nao
Comment by Julie White posted on
Hi Tom,
As this blog has recently been highlighted within the recent 'The year’s most popular posts' round-up, I wondered if there are any plans to update your blog with the last 12 months usage stats?
I'd be really interested to see if these trends have continued, or indeed accelerated?
Comment by Adam Branscomb posted on
I love this detailed analysis, and the fact that you're not just taking a snapshot but monitoring trends. On our UK local authority SaaS product, we are also seeing IE6 dropping to insignificant levels, but we still (in January 2014) get just over 4% on IE7, rather than your 1.5%. I suspect this might be because a higher % of our traffic is essentially intranet - internal to the local authorities. I'll have to see if we can differentiate this in our analysis.
So the next question is, at what point will this analysis drive changes to GDS guidance? Specifically the Government Service Design Manual at https://www.gov.uk/service-manual/user-centered-design/browsers-and-devices.html. I suspect (and hope!) this gets an update soon...
Comment by Nathan Wall posted on
Hi Tom
This is really useful data for us to compare with what we're seeing here in NZ.
I'm happy to say that our IE6 use is only 0.06% - which does put a smile on my face 🙂
Chrome is only marginally ahead of IE as the preferred browser, 32.8% : 31.1%
Mobile use of our site is just under 18%, and that's roughly what we'd expect based on trends we've seen on other NZ government sites at the moment. It will be interesting to see if that increases once the site comes out of beta, the current version of the site doesn't support smartphone display at all and that's probably putting some users off...
Keep the data coming...
Comment by Charles Meaden posted on
Would love to see the brower resolution sizes split by desktop, table and mobile. I can make an educated guess as to which is which.
Comment by Floyd Price posted on
Reading that IE6 usage has halved is music to my ears and i can't help feeling a collective "Smile" within our industry when I see statements like that.
With that said I'd like to see less hostility towards IE in general within our community, version 9, 10 and 11 (and beyond) are perfectly capable browsers and dare i say it, "Standards Based".... well almost but the landscape is starting to clear which I think we can all agree is fantastic.
Please keep the data coming, more granular breakdown would be appreciated.
Comment by JF posted on
I'm slightly confused by the numbers?
The bit about OS & browsers says: Sample size Jan 2014: 36.4m visits
But the bit about devices says: January 2014. Sample size 47m.
Comment by Tom Loosemore posted on
Sorry - my fault for not being clear enough. The original blog post was written on 22 Jan, and the OS, browser and screen res tables use data for the month up to and including Jan 20th (ie including xmas dip). This gives 36.4m visits. The devices data I added yesterday was for the calendar month of Jan 2014, giving 47m visits.
Comment by Tim posted on
Interesting, but obvious.
It's odd albeit trendy to lump all versions of iPhone together and split Samsung Galaxy to make Apple look better.
Of course mobile usage is growing, because they've just rebuilt and redesigned all of these government websites and the old ones were terrible and generally unusable on mobile! That's a major reason why they did it. It's a trendy self-fulfilling prophecy.
The 10% difference in desktop market share risks stating the obvious: people are using mobile as well as desktop these days. But market share is a misleading stat in isolation -- e.g. share can fall even as usage rises. The big PC screens are gaining popularity.
It's such a trendy article it will be widely shared and cited. It's ideal for Apple-obsessed Bootup which indeed is where I found it.
Comment by Tom Loosemore posted on
Hi Tim,
Thanks for your comment. We'd very much like to be able to break out iPhone and iPad data into more specific iOS and device versions, but we can't get at that level of detail at the moment. We've published the most granular data we possess.
It's a valid point that the usage of mobile increased in Oct 2012 when the responsive GOV.UK replaced the previous stand alone mobile sites - as the previous blog posts say, our mobile traffic doubled from 10% to 20% overnight on the day GOV.UK launched. However, the growth in mobile and tablet use since then has also been very significant, and government services must respond accordingly.
-Tom
Comment by Scott posted on
Is there any chance of a breakdown of IE versions? I'd like to know IE7 usage is negligible and IE8 is disappearing... 😉
Comment by Tom Loosemore posted on
Hi Scott,
Still want to get post published wuth proper data and pretty graphs for browser versions. Quick peek says IE7 was at 1.5% in Jan 2014, down from 2.8% in Jan 2013. IE8 still at 9% in Jan 2014, down from 13% in Jan 2014.
HTH
-Tom
Comment by Chris posted on
Is there anyway to know what the breakdown of form factors are?
For example is the iOS usage primarily phone or tablet based. Likewise for Android.
Comment by Tom Loosemore posted on
Hi Chris,
I've updated the post to include a top 100 devices used in GOV.UK in Jan 2014, which will hopefully help. iPhone ahead of iPad by a bit.
-Tom
Comment by Graham posted on
Hi Tom,
Thanks for sharing this - made me dig into wales.gov.uk analytics! Found a similar picture, although with a few nuances:
* IE usage is not dropping as fast as GOV.UK - it continues to dominate with twice as many users as Chrome (although IE6 has more than halved in 12 months)
* we're seeing a slower drop in Windows (from a higher start) and a faster rise in iOS (from a lower start)
* we've seen a 7% rise in 1366x768 compared to the 8% drop on GOV.UK - the only finding that wales.gov.uk is going in the opposite direction to GOV.UK
Comment by Stephen Edwards posted on
We should hold a party to celebrate the demise of IE6 soon ... when it gets below 0.1%?
Comment by Alex Bilbie posted on
It'd be great if you could render this data out into monthly CSV dumps and upload them to data.gov.uk 🙂
Comment by Tom Loosemore posted on
Hi Alex,
The performance platform team who run http://gov.uk/performance are very keen to expose GOV.UK data as data. I know they're working through the best ways of doing so sustainably.
-Tom
Comment by Chris Brandrick posted on
+1 for this.
Comment by Ross Chapman posted on
Interesting to see you solely using Google Analytics - do you use it pretty much out of the box, or have custom events/routing setup?
Thanks for the post!
Comment by Peter Jordan posted on
GOV.UK is currently instrumented with Google Analytics.
Absolute metric values are not actionable and so we've used a mix of out of the box features and custom variables to build segments and comparative views. Understanding user behaviour trends over time and by different segments; such as parts of the site, different content formats (such as guides or smart answers), referring channels, devices or browsers helps product managers to iterate the many formats and elements of GOV.UK.
It's always a moving feast as analytics products improve and GOV.UK itself develops.
Comment by Mike Thacker posted on
Thanks very much. Really useful data for all of us developing for a similar user base. And nice to see IE6 declining.