I was reading a news story the other day about the iPad, and how this totally new device was changing everything about the way people were accessing content. Revolutionary as the iPad may be, as a tablet user of many years now I tend to take this kind of story with a pinch of salt, but it did make me wonder whether or not I could do some digging into web stats to see just how much other tablets may or may not have been used over the years. Trouble is identifying a mobile device within web analytics software, like Google Analytics for example, has only become a simple operation in recent months, so historical digging was going to be difficult.
It occurred to me though that there is one very fundamental way of knowing whether or not someone is using a tablet device. Most tablet computers – and for that matter mobiles – are designed so that they can be used in portrait mode as well as landscape mode. All I would therefore need to do is look back in the data to find instances where the horizontal resolution recorded by the web analytics software was shorter than the vertical resolution – not perfect by any means, as tablets can be used either way (and for that matter some desktop users prefer a veritcal display) but good enough for a useful understanding I thought. Little did I know that I was about to uncover more than I’d bargained for!
As this was primarily about the iPad impact and tablets, I first wanted to find a timeframe when the iPad wasn’t available. I also wanted to use a timeframe long enough to generalise from, so plumped for a sample from 1st April 2009 to 31st March 2010. My first shock was the sheer number of different resolutions. Over the year that I’d chosen 2,170 different screen resolutions were recorded as having accessed our websites. 2,170! I expected maybe a couple of hundred, not a figure in the thousands. This was obviously going to take more work than I’d anticipated. The second shock was the resolution at the top of that list, and not just at the top of that list but at the top of that list by far – 1280×800. Here’s the breakdown for the top 10 of those 2,170 resolutions.
Now if you’re not too familiar with this field you may be wondering what I’m on about, and why this is important, well the reason why I was surprised is this top number is a laptop resolution, not a desktop resolution. It shows that the vast majority of our web traffic comes from people on laptops, not desktops.
On reflection you only have to wander around campus to realise just how many students, for example, are accessing the web using their laptops, and of course this is going on behind close doors much much more than it is in public, but I don’t think I was expecting the discrepancy to be quite so large. Further analysis of the data for the top 500 resolutions showed that very nearly 70% of all access to our websites is through a widescreen device (i.e. greater or equal to a 16:9 screen resolution), the majority of which I think it is safe to say would be laptops.
But back to the reason I started this blog post – just what can we learn from web stats about screen resolution and peoples browsing habits with tablet devices? Well, interestingly, about 185,000 vists to our sites were done in portrait mode, suggesting that people were using either a tablet or mobile device to browse, but that’s only 0.66% of the total number of visits. If I take it a step further and exclude small screen devices (so only including devices with a horizontal resolution whilst in portrait mode of greater than 480 pixels) the figure is much, much smaller. Then it seems only 0.03% of vists are returned.
So the conclusion of this experiment? Well it seems that the there has been very, very little use of tablet devices so far at Exeter, at least in portrait mode, but that may well change in 2012 as every developer you can think of seems keen to jump on the iPads coat tails and join the tablet PC market. But what I think has been much more interesting is to appreciate just how many different types of device are out there trying to connect to our websites, and for that matter that the vast majority of these are almost certainly laptops. Whilst I think it’s fair to say that there will always be a place for the desktop computer, and the computer suite, at Universitycampuses such as ours, how exactly we design our IT infrastructure and in particular the amount of effort we spend on the provision of different services may need to shift in order to take account of the changing nature of access.
Perhaps it’s time for the desktop development units and the web development units to start working more hand in hand, and for some of the – perhaps artificial – barriers between IT disciplines to be dismantled. The desktop of the future is almost certainly going to be a web based desktop, and for mattter a web based desktop that will need to be accessible – and usable – on a multitude of different screens, and what looks like a huge variety of different screen resolutions. By moving our infratsructures together in such as way to anticipate this change before it happens we will be much better placed to deliver the world class service we need to both our staff and students alike.
There is a difference in the ‘way’ people access web content now using touchscreen devices, which is revolutionising web interface design. Your extensive experience using a tablet is based on single point pen input, not multi-touch. The ‘way’ is most important, not the ‘how’.
It is also clear that widescreen is the predominant screen ratio (although the iPad is actually 16:10. I would expect 1280×800 to be at the top of the list), however I don’t know if there is a tablet available that uses this resolution, were you attempting to “see just how much other tablets may or may not have been used over the years”, doesn’t this mean that laptops should be excluded from the stats if you are looking at tablet stats?
Are there any other indicators that could be used to identify the difference between mouse/touch/pen input?
These are important figures, showing that we really need to think about web design in both portrait and landscape, as devices easily rotate nowadays. There is no wrong or right way to hold it, after all. However, I almost always use my iPad in landscape mode for web browsing at 1024×768.
Very interesting article, more research needed.
Really interesting article. I remember my first HTML pages back in the day were always designed for 800*600. Guess most of them would look a bit, well ‘sparse’ on a vast majority of today’s screens judging by the stats.
Interesting analysis about students’ use of laptops. When I was a student (05-09) I don’t think any of my friends had desktops at all, so I would suspect a 95%+ use of widescreen resolutions after excluding visits from staff and cluster pcs.
One small caveat tho – widescreen ratios seem to be increasingly common on desktop pc monitors too nowadays, so perhaps it’s becoming less easy to distinguish desktops from laptops in the first place???