Say Goodbye to Done
by Rachel HinmanI found this recent paper published by the Nokia Research Center (via Putting People First) an extremely interesting read. It echoes and builds on some thoughts I’ve been having recently about mobile user experience and interaction design. What has long excited and inspired me about mobile user experience is how it presents the opportunity to explore new ways for people to interact with information. The Nokia paper depicts a future where people and their mobile devices will be part of a self-organizing ecosystem of data. It’s exciting stuff, but the question remains for me… How do we begin to design for that future?
One mobile user experience trend I’ve been tracking is the slow erosion of a task-based interaction model. Most software, web sites and web-based products we use today have evolved around the task-based model, and it has served us well. PCs are great tools for efficiency and “getting stuff done”. Designers are well armed with a vast set of tools and processes that support this approach – use cases, task flows, task analysis – just to name a few.
The thing is… mobile isn’t a great platform for accomplishing tasks. The small screen and variability of the mobile context leaves most users feeling lost in a labyrinth of menus.
If PCs are great for getting stuff done, mobiles are good at exposing possibilities. More and more, I’ve been thinking that to create great mobile experiences, designers need to say goodbye to tasks, say goodbye to done… and explore new or different interaction models that leverage the things that mobile is good at. Exposing possibilities.
Here are three emergent interaction models that I think support the idea of exposing possibilities in the mobile context:
1. Accrue value over time
2. Facilitate exploration
3. Sense intent
Interactions that accrue value over time
This interaction model shifts the focus from task completion to surfacing information and making it easy for people to participate. A great example is Twitter. I’ve long heard folks who’ve never used Twitter ask, “What’s the point?” Compared to a similar experience that uses a more task-centric model like email, Twitter’s value is only revealed as users engage with the service over time. The value of the interaction is not around completing a task – typing a response to “what are you doing?” – but rather the conversation that can happen as a result.
Interactions that facilitate exploration
This is an interaction model that calls to mind two of my favorite iPhone apps – Koi Pond and Bloom. These are open-ended interaction models that are easy to enter and exit. The interfaces usually have built-in affordances that inspire curiosity and play. They usually have some type of clear and immediate feedback, are visually rich and engaging, rely on animation to aid in cognition, and often orchestrate touch, gesture and sound into the experience. Pointless? Perhaps. However, there is something so completely intriguing and fun about these interfaces that is far more emotionally satisfying than clicking a send or buy button on a web site.
Interactions that sense intent
This interaction model is one I’ve been tracking for the last 18 months and is perhaps the most exciting of the three. This model uses information from sensors, use patterns, GPS data and algorithms to anticipate needs and deliver intuitive options that make sense in a particular context. Devices are already doing this today. Sensors and accelerometer data on the iPhone can sense the orientation of the device and adjust the interface and screen orientation accordingly. The mobile Google Maps application anticipates that users will want to use their current location and automatically integrates it into the interaction. This model seems to be less about enabling users to complete discreet tasks and more about sensing what users want and delivering intuitive options.
I doubt tasks will ever be banished from our mindset completely and they shouldn’t be. The task-based model has been a good friend that’s served us well. However, it feels like the only way we can realize the opportunities that mobile interaction design presents is to say adios to our old friend, “the task” for a while and focus on making some new friends.
Entries (RSS)
January 19th, 2009 at 1:01 am
A thought provoking article Rachel. I too would like to see wider exploration of more varied interaction models, particularly using the growing range of sensors and contextual feeds we now have access to. I think there is an opportunity to appeal to the playful nature of the human character in an interface, where users participate because to do so gives them pleasure rather than simply allowing them to tick a task of their mental list – Nintendo has been great at pioneering this kind of design approach.
However, part of me struggles to see how we will overcome generations of task-based thinking among humans, at least in many Western cultures. We’ve been brought up on the idea that to complete as many tasks as possible, as efficiently as possible, is the path to achievement. This mentality is ingrained, certainly in previous generations, and satisfaction comes from seeing tasks ticked off. Perhaps the generation growing up today will see things differently, influenced by the constant nature of IM, Twitter and other digital interactions, where these engagements are seen as ongoing commitments rather than discreet transactions?
January 19th, 2009 at 1:31 am
A thought provoking post Rachel. I too would like to see greater exploration of interfaces which encourage humans to interact with people, content or the world around them because to do so gives them pleasure, rather than just allowing them to check a task off their list.
I believe Nintendo, for instance, has made some positives steps in this direction by tapping into the playful nature of the human spirit in all ages, allowing it to break outside the core 14 – 24 gamer group. Witness Wii parties among older generations and the 70 year old lady sitting next to me on a recent flight engrossed in her Nintendo DS!
However, I am also concerned task-based thinking is deeply ingrained in human nature, at least in Western cultures. We have been brought up in an education system which values task completion and where achievement in life is often defined by who is able to complete defined tasks as efficiently as possible.
As a result, we look at many of our day-to-day interactions as discreet transactions, each with a start and end point. Satisfaction is derived from being able to draw a line under each action and marking them ‘complete’ in our mental check list. This is true in both the digital and physical realms, at least for those over the age of 25, who have grown up in a world where computers are something that has arrived rather than something that has always been there.
Perhaps the generation growing up now will be the first to have a different approach? Tools like IM and Twitter are seen rather as an ongoing conversation or commitment than a transactional task. Those who have been brought up on them may be more comfortable with the idea of explorations and interactions which continue indefinitely than previous generations.
January 19th, 2009 at 8:34 am
This is a really insightful article, and from the perspective of a heavy mobile user, one that makes me think even more about how I want to “get things done” on my mobile and how that mindset really just doesn’t work in many of the usage contexts.
I do think that you have a great point here, and that there are some devices and applications that make this happen easier than others. But it would take a near act of God to get a whole-scale change of thought on the side of many users who are just too used to that PC/task-based paradigm.
Having said that, I do think that mobiles should be able to have multiple interfaces per device, based on the context of use. Not just in the sense of application on top of UI, but system-level UI changes based on the orientation and context of the device. For example, sense that the mobile is moving, and move to a voice-activated, large-icon view (thinking that one is in a car); or, device plugged into TV and the resolution and interface changes accordingly.
Great thoughts, and thanks.
January 20th, 2009 at 1:39 pm
Great article, and I like the categories.
I’ve been thinking about intent-based approaches for situated applications as well, and find it helps us think more about enhancing a users context, rather than just providing cool features.
I think one of the key points is that, prior, we tried to predict intent by analyzing past behaviors – with all the attendant problems that implies. With mobiles now adding sensors like orientation, accelerometers and GPS, we can add real-time data into the analytical mix.
Tough to do, but we find that even just a little consideration of user intent adds a lot of value.
thanks,
::t
January 20th, 2009 at 4:53 pm
Thank you for this article Rachel. I guess that the nature of our jobs obliges us to simplify things a little.
However, by simplifying, there is a big risk of abstracting some notions and unfortunately when we abstract basic notions we create confusion.
The three interaction models you exposed are genuine but I must say that they have nothing to do with the exclusion of tasks from the model. In my opinion, there is a prominent confusion between “Purpose” and “Task” in this article.
The answer to the question “What’s the point?” is not Task but rather intent and purpose.
Tasks translate concretely into one or more functionalities. Possibilities are another way of saying functionalities. If tasks lead to functionalities and functionalities are possiblities, then tasks lead to possibilities. Now, the beauty of your article is that it suggests possiblity-based models…But let’s think practically, how will you translate possiblities? – By defining tasks!
Let me illustrate this with one of your examples. One of the possibilities you mentionned is the interactions that accrue value over time. How do you elaborate on that? – by saying “we switch the interaction model from task completion to surfacing information and making it easy for people to participate”…Correct me if I’m wrong, but isn’t Participation a task that could be divided into sub-tasks in order to understand the process, simplify it and make it easily accessible to the user?
So the parameter Rachel that needs to be reconsidered isn’t, according to my understanding, the Task but rather the intent and the purpose of use of the product. Because a product might have no purpose at all but that doesn’t mean the users won’t make-up and set their own objectives, and the least of those is pleasure…
You have all seen the ZIPPO application on the iPhone… What’s the point? – Nothing! (Purpose = nothing)What can happen? – A lot! How do you use it? – that is where Task becomes important because if it isn’t for a task-based model, good luck trying to light it up!
Let us not confuse strategy with design. Strategically speaking, possibility-based models are the way to go, as they open up horizons for multiple uses. However when it gets to design, possibilities need to be elaborated. Then, we’re no longer in a possibility-based but in a task-based model. A good start for strategy is possibilities but if we are to design for great interactions, we have to understand the process and the best way to understand the process is to identify the task.
Let’s remember that all those great interfaces you talked about came out in the era of task-based models. I’m not sure their creators even thought of the paradigms you’re suggesting in this article.
January 20th, 2009 at 10:09 pm
I guess we will have to agree to disagree, Dory.
True, one could approach any interaction design problem from a task-based mindset.
Just as one could look at life as a series of tasks to be completed.
It’s very logical and rational – and it certainly would be less confusing.
However, if you think of the range of human experiences, there’s actually a lot of them that would seem awkward at best if viewed through through the “task” perspective. Most human relationships like marriage, parenting, friendship… or activities like dancing, flirting, joking, and dating could be broken down into a series of tasks and sub-tasks to be completed by “users”. But in doing so we lose the subtlety, grace, joy… and frankly the humanness that makes these experiences special.
Just as there is more to life than tasks, I believe there is more to interaction design than task-based models.
If the goal is to make technology experiences empathetic to the human experience, why do we constantly turn to the task-based efficiency interaction model for most technology solutions? What are we losing? More importantly… what are people who use the stuff we design losing?
I think the time has come – especially in the realm of mobile user experience – for the work we create to reflect these more human qualities. To do so, I argue that we need to be capable of looking at the methods and processes we currently use and acknowledge them for what they provide *and* where they fall short.
If we suspend our belief that tasks are the panacea, think of all the other types of interactions we could create…
January 21st, 2009 at 6:56 am
I totally agree with the humaness of interaction design and everything you mentionned in this last email. I am very enthusiastic to see how you are going to take it from there and especially how you are going to translate this model practically… as long as it doesn’t pile up with the other utopian theories and philosophies.
Thank you again for this article.
January 21st, 2009 at 11:28 am
Count me a skeptic in sensing intent. The iPhone/iPod Touch interface often lets me down based upon the two examples you cite:
Am I the only guy in the world who’s opened a web page, flopped down on my side in bed and had mobile Safari rotate the screen so I have to read the page sideways?
I can’t have been the only one to try and have a discussion about a Google map of an area hundreds of miles from my current location and struggle to get the mobile Google map app to let go of my current position on the map.
Douglas Adams wrote a lot about the mythical “Sirius Cybernetics Corporation” and their complete failure to produce robotic products that please their users. What Adams captured was that the products often got user intents completely wrong, and delivered just the wrong thing at the wrong time.
Unless your mobile can know – really know – your intent, it’s going to produce some frustrating behavior for some people some of the time.
So how can the mobile know my intent?
In time-tested interaction models, the user presses buttons or sets preferences.
In an iPhone context, why not allow the user’s gestures to override the so-called “predictive” ability. Let me twist the screen back to where I want it and have it stick that way until the next twist or power-off event. Let me swipe the current position right off the map.
It might be helpful for the mobile to guess at my intent sometimes, but when the mobile has gotten it wrong, there’s nothing more maddening than not offering me the option to straighten the mess out.
January 21st, 2009 at 8:02 pm
As the paper highlights, the combination of sensors, mobile phones, and people will create new opportunities for future services through real-time data collection. And that is really exciting for designers. What that has to do with tasks, I’m not sure. And how mobile devices are not good for accomplishing tasks is also unclear to me given all the tasks you can complete with a mobile phone that you cannot on a computer. Certainly, mobile devices offer different opportunities for experience when compared with desktop computers. They have different constraints and, as the article suggests, advantages. However, one is not more or less task capable than the other.
Defining tasks are for good describing how specific parts of a system should work. Tasks are part of an experience, where the experience is the whole. A good designer needs to understand the relationship and design for both the parts and the whole. Twitter would be nothing if people could not complete the task of updating status. But Twitter is not just the sum of its parts. That’s why in addition to task flows, designers use other methods to communicate the intended experience.
Understanding the use of a product or service over time, the level of playfulness or exploration it might have, and intent of its use across interactions seem like things designers have been considering for quite a while. Were these emerging interaction models taken from somewhere? It is my hope they are not emergent at all, but considerations for the design of any product or service.