XUX: Cross Device User Experience
With the growth of the mobile internet, the rise of crossmedia productions (Dutch crossmedia blog), and the growth of cloud based applications (my post about cloud computing) interaction designers and usability engineers face a fairly new problem. The difficulty is to deliver good usability and user experience for the service across multiple applications running on multiple devices and different contexts of use. I can access my e-mail, microblogging, blogging and social networking services from the desktop and my mobile phone. But is the use of these tools in one context affecting their use in other contexts? I believe there are some tricky new challenges to cross device user experience design. In this post I would like to sketch a few of them.
It seems like managing Users Experience and usability across devices has received only cursory attention from academic researchers. Whiley™ published a book on the topic edited by Seffah and Javahery, but the book dates from 2004 when cross device usability was still a theoretical rather than a practical problem. IBM researchers Pierce and Nichols explore the issue in a UIST paper in 2008, but their focus is mostly technical. The Nielsen Norman Group studied the usability of the ‘mobile web’, and found it hardly usable. Jacob Nielsen compares the mobile web of today with the internet of the early 90ies, an apt comparison, because the mobile web is about the same age now as the non-mobile web back then. Finally, there is a good Interact paper by Microsoft Research and the University of Washington.
Should it matter on what platform you are using a website? Wasn’t the web created with a design for all philosophy in the first place? When I first learned about making websites, we learned to make sites in such a way they were accessible for all clients on any conceivable platform, even text to speech devices. In practice we never took the advice, nor did anyone else (apart from Nielsen himself). Honestly, the average effort to make the site ‘device independent’ limited itself to making it accessible for those poor people with only 800*600 pixels on the screen. It turns out that now, 10 years later, Jacob Nielsen’s own website (UseIt) isn’t suited for 320 by 480 pixels of my iPhone, but it still stands out as very usable compared with others. Most sites are much, much, worse.
To be honest, I am not a believer of device independent interaction design. Surely it should not make a difference whether I am using Windows or Linux as an operating system, for the usability of a web site. But the hardware I own and my context of use, should make a difference. If I have multitouch possibilities on my device, and I am doing more with the device than reading blog posts, I want the applet to make good use of the multitouch. When the interaction space and context of use are different for different devices, device independent design means settling for the worst solution. Rather I would like cloud services and cross media providers to deliver good device specific UX, for most of the important platforms that they target. But this has its own challenges. I will mention two.
1 Deciding on Functionality
My iPhone app for WordPress has only limited functionality. I can check and edit comments and can read and write posts and pages. Personally I am not much of a mobile writer, so I would never write a blog post, or reaction to a post from my mobile. I am a statistics junkie so I miss the possibility to dig into to those from my mobile phone. In fact I don’t use the WordPress app because of this mismatch. Still, beyond the specific choices WordPress happened to make, you can ask the general question whether mobile apps for on-line services should have a different functionality at all.
There are two reasons to pick a subset of the functionality in the mobile case. First, because of different hardware constraints such as a small screen and a virtual keyboard, it may be hard to deliver the full functionality in a usable way. Second, mobile use may have different (dominant) use cases. For example, delivering a web experience on a mobile phone, entails a focus on the ‘real-time web’. What is ‘new’ and what users would want to do with this news during a lost moment in a train, between meetings or during (boring) meetings. More time demanding tasks such as the composition of a new blog post are not a focus in a mobile application.
The difficulty with limiting functionality, is that users build a model of what they can do with a service across devices. I would like to have stats in the WordPress app but others might support for different tasks. Few people may want to write blog posts from their mobile, but omitting the functionality to gain usability will offend those users. They know the service can do it, and so they want access to this functionality from their mobile.
There are a couple of ways out of this dilemma. First you can ‘hide’ functionality that supports non-dominant use cases behind shortcuts, or put them ‘deeper’ in the interface. But this may hinder learnability. Second you can make a configurable interface that can be extended with modules for certain functionality. This is a complex UX problem in itself, but in a way this is what is happening on the iPhone in the main menu. Third, you can ‘outsource’ the problem by publishing an API and allow third-party developers to create apps for your service that meet special needs. In a way this is how Twitter solves the problem. Of course this means you need to mobilize developers and it puts a load on the user to pick and learn the different applications.
2 Portability and cross device tasks
Even trickier than deciding on a functional specification is the issue of what I would like to call portability. Sometimes a friend shares an interesting blog post in a Tweet, I click on it on my mobile read it in the native browser of my mobile phone and decide to bookmark it. I cannot store this as a bookmark of the local browser, so I send the bookmark to my computer by e-mail after which I store as a bookmark. According to the researchers behind the Interact paper, this is what many people do. Of course, I could use a cloud based (social) bookmark service, which could provide a plugin for my mobile browser, so my mobile bookmarks carry over to any device I would like to use. But the underlying usability problem is more general. As a user: how do I know whether a change I make to the system, a preference, a change of view, a text edit effects changes in another view. And how do I control this? If I resize my text on the mobile, will it change in my web browser? It may seem trivial that it doesn’t, for designers but it is not for users. Moreover, the problem is not solved consistently so users do not have a way to build a good conceptual model of the service.
A related problem is that tasks may extend across devices. Broadly speaking, bookmarking is a good example. I might scan a text, when I am on the move, and post them for later reading when I am at work, or vice versa. I might do editing of a post on one platform, and write it on another. This brings portability challenges, because some changes should be local, some should be global. With collaborative work, there is a social dimension as well. How would Google Wave work across devices?
As applications move into the cloud and cross media solutions will become more common, UX designers get new usability problems to work with. Of course the cross media usability will depend on the usability of the independent interfaces but it is not a compositional problem. On top of the usability of the independent platforms there are new problems to consider in this post I mentioned two of them. I must say that I had a hard time to sort out the difficulties and framing them legibly. Let me know if you can help out.
Filed under: (re)thinking media, probe | 2 Comments
Tags: Cloud Computing, Cross Device UX, Crossmedia, Holistic Web, iPhone, Mobile Web, Usability, User Experience