I live and breathe user experience design, and yet it took me two years to get myself the device referenced by almost every single presentation about user experience since 2007… Apple’s iPhone. My reasons were very specific and perhaps boring, but what is interesting is the perspective this wait has afforded me. Since it was released, the iPhone has grabbed an astonishing share of mobile Web traffic, been regarded as a “game-changer” in both the design and business worlds, and has even been referred to as the “Jesus Phone.” Now that I’ve owned one for two weeks I’ve developed a different perspective. The iPhone is surprisingly difficult to use, but it sure is fun! And that is why it’s a game-changer.
A Lack of Affordances Leads to Low Learnability
Learnability contributes greatly to the usability of a system. If a system is designed for a specific context, it should be easy for people in that context to approach it, assess its controls, and manipulate it. Granted, learnability isn’t everything, but when it’s tough to figure out how to do things you’re on the express bus to a frustrating experience. There are two things about the iPhone that contribute to its difficult learnability. It lacks physical affordances and suffers from inconsistent visual cues.
Gestural user interfaces (UIs) are the 21st century’s version of the command line interface… they’re really fast and easy provided you’ve memorized a bunch of commands. This is fine for those who are accustomed or inclined to explore a device, but many people just want to check their calendar, write an email, or make a grocery list. These people will react to what they see on the screen rather than explore possibilities, which leaves them out of luck with a gestural UI.
The iPhone’s featureless touchscreen is Don Norman’s proverbial glass door. Apple has done a stunning job of making things that are pressable look like they’re pressable, but that will never be as effective as an actual button. With physical, simple buttons we can rely on motor memory to manipulate a device without paying attention. But the iPhone’s buttons are highly contextual, which forces us to pay attention to the device to remain aware of its context even after extensive use of the system. The “problem” is that the iPhone is a convergent device, a device with multiple functions. With 50,000 apps, you might even say infinite functions. The only way to build a device that serves 50,000 different purposes is to make it almost entirely free of physical affordances. Of course, the big value proposition of the iPhone is that it is the first mobile device to achieve an effective convergence.
Pressing a button is an action that a gestural UI can communicate visually, but there are a number of other actions that have no visual cue. Direct manipulation gestures such as tap (on something other than a button), double-tap, tap-and-hold, swipe, and pinch/zoom are far more difficult to communicate. These rely on user experimentation and memory.
Even worse are the modal gestures such as shake to undo and swipe to delete. If users discover them at all it’s usually by accident. They don’t map to anything (outside of an Etch-a-Sketch) and there are no clues to indicate that they’re available. Being mentioned in a WWDC keynote does not count as a clue.
Inconsistent Visual Cues Don’t Help Either
Apple has gone to great lengths to make the UI consistent, even publishing the iPhone Human Interface Guidelines, but some inconsistencies remain. Application buttons can have labels or not. Some applications, like TweetDeck, AP Mobile, and others, obligingly label their buttons:
Others, mostly Apple applications, do not:
Does the circular arrow mean reload like in Safari? Or reply? If it means reply, what does the other arrow mean? Labeled buttons communicate their functionality much more clearly. (The circular arrow does mean reload, but makes no sense in the context of a message. The swoosh arrow does mean reply.)
The landscape keyboard, despite being a basic device function, isn’t supported by all applications. When it is supported, there are no visual or other cues that indicate it. Not only is it difficult to learn when the landscape keyboard is available, cues as to its availability are stored in only one place, user memory.
Even the iPhone’s implementation of its standard gestural interactions is inconsistent. This is most frustrating on simple interactions like tap. There are obvious tap targets like buttons and non-obvious targets like received calls, tweets, emails, etc. In some cases, a tap on a non-obvious target means “open” or “get detailed info.” But in others it means “take action.” The worst example of this is the Recent Calls list. I can’t tell you how many times I’ve accidentally called someone when what I wanted to do was get more details about the call. Yes there is an arrow button, but it’s on the right side away from my focus. Other applications (like Mail) have trained me to tap an object to get a detailed view of it, so my natural tendency is to tap the contact name or number.
When applications do not implement buttons, device functions, and non-obvious gestural interactions consistently, this increases the learnability problem. Not only do users have to learn and memorize what the device does, they have to learn how each application makes use of those functions! This is much less of an issue in point-and-click interfaces, which require fewer physical interactions and present most options on the screen for users to react to.
If the iPhone is so difficult to use, why is it still regarded as a game changer by both the design and business worlds? Because it does several important things right, but most of all because it’s fun.
Fun is the New Usable
As a user experience designer, I thought my job was to make things not suck. Until recently. As technology has evolved, human behavior has evolved along with it. Since behavior is the basis of user experience design, my job has evolved as well. Now, my job is to make things people love. At the 2009 IA Summit, Karl Fast articulated the value proposition of user experience design with sparkling clarity. “Engineers make things,” he said, “we make people love them.” And then he held up an iPhone as an example.
This is a crucial change, the importance of which cannot be overstated.
Any new system or gadget has a learning curve, but where the iPhone differs is that the nature of traversing that curve is more fun than frustrating. You swipe and pinch and tap and shake your way to familiarity instead of pressing awkward buttons and navigating byzantine menu structures. You learn the iPhone by playing with it, which encourages interaction because humans are built to play. Even in a system like this, we could quickly be dissuaded from doing so if wrong actions had negative consequences, such as getting online or sending messages accidentally. The iPhone is mostly devoid of these sorts of consequences. The only time I’ve run into this is repeatedly calling people I didn’t want to call while viewing my Recent Calls list.
The iPhone goes further than encouraging play; it rewards play. If you explore the phone’s applications, you will often find them anticipating your needs. When viewing a video you’ve shot and press the action button, you can email it or upload it to YouTube. If you try to email it and the video is too large, it will ask if you want to send a smaller clip from the video instead of preventing you from sending it. The iPhone then presents you with the UI to trim a clip and continue with your message. The original video remains untouched. Simple, sensible, satisfying.
Effective & Delightful Convergence
On the day I got my phone, someone sent me an email that contained a physical address. The phone turned it into a link. I clicked it, got a map, and the phone asked me if I wanted directions. From my current location. I giggled excitedly.
The delight induced by how well the iPhone’s applications interact with each other is another reason for its success. This is the point at which usability and playfulness intersect. The experience of having needs not just met but anticipated creates the joy that encourages users to continue exploring. This intelligent interaction between applications is absolutely key to making a convergent device delightfully convergent.
But you can’t have a delightfully convergent device that isn’t effectively convergent. What converges are contexts of use. The interactions between applications that I described above represent relatively minor, detailed contextual shifts. These small shifts result in delight only if the device handles major shifts effectively as well.
Before I had an iPhone, I would switch major contexts by switching devices. If I got a call while listening to my iPod I’d stop it, put it down, and pick up my phone. This was intuitive to the point of being instinctual. But now my iPhone must handle that switch of context for me. If it failed to do that in a sensible way, I would think the iPhone sucks. An effectively convergent device is one that, like the iPhone, can handle major shifts in context in a way that supports the user’s transition between those contexts.
Implications for User Experience Designers
iPhones fly off the shelves despite being difficult to learn.
Because they let you do what it says you can do and they make you happy while you do it. This proves that my job as a user experience designer has evolved rather than simply changed. While it’s still my responsibility to prevent things from sucking, now it’s also my responsibility to add a little playfulness. As Kim Goodwin said in her Interaction09 keynote, we have a limited window in which to prove how valuable design can be to business. There are three ways in which user experience designers can learn to incorporate play into the systems they design.
Experience and Research Play
You can’t build playfulness into your designs without experiencing playfulness yourself. Play games and pay attention to what makes them fun. For example, the only rule in the card game Fluxx is that the rules constantly change. Completing a level in Peggle gives you the “Ode to Joy,” rainbows, unicorns, and fireworks! Use these elements as inspiration for working playfulness into your designs. You might not be able to play the “Ode to Joy” when people complete a purchase, but can you delight them in another way?
Play is a behavior. As a user experience designer, you should explore research about play and playfulness just as you’d explore research about gestalt perception or information seeking. The National Institute for Play is a good place to start. The ACM digital library has some resources on playfulness and computers. Questia has resources on play in general. (Both ACM and Questia are paid services.)
Become Familiar With Game Design
Game designers put a lot of thought into how to design a fun experience. We can learn a lot from the principles they use to make this happen. Much of game design seems to revolve around creating, sustaining, and developing a narrative. This aspect is less important to user experience designers than game mechanics and the design of casual games.
A game mechanic is anything that guides the play of a game. Most mechanics take the form of either rules or possible actions. In cribbage, players must discard two cards to the crib (rule) and they keep track of their progress by placing pegs on a board (possible action). Game mechanics translate into the user experience design world as interaction patterns. Understanding how game designers make games fun by designing pleasing game mechanics will help you design pleasing interactions. The Critical Gaming Network’s Game Design 101 has a great discussion of game mechanics.
Casual games are those that are meant to be picked up and played simply for the joy of playing them. They are always enjoyable, often compelling, but not engrossing. Peggle and Paper Toss are canonical examples of casual games. Casual game design is important to user experience designers because they place special emphasis on learnability and delightful interactions. When we design systems that are fun, delightfulness should be a side effect of interacting with them even though it is not the goal. People still have tasks to complete and we can’t let fun get in the way of that. For more on casual games, read the Casual Game Design blog as well as Nicole Lazarro’s “Why We Play Games.”
Re-Learn the Art of the Tutorial
My experience with the iPhone has led me to think that maybe fun doesn’t need to be intuitive. Maybe fun is so valuable that people will make the effort to learn a system built on fun interaction patterns. If I had the opportunity to change one thing about the iPhone, I would add a tutorial. By tutorial I don’t mean a boring list of stuff you can do with it. I’m specifically thinking of some sort of mini-game. It would introduce users to all the different gestures they can do and the contexts in which they’re appropriate, challenging them to choose and perform the right one.
Tutorials in the casual games I’ve played take one of two forms. The first is much like the mini-game I described above. This type of tutorial is composed of levels in which the goal is to learn, explore, and practice one or more game mechanics. The player then begins the “real” game. The iPhone games Isotope and TaxiBall contain good examples of this type of tutorial.
The second is the in-game tutorial. In games with this type of tutorial you simply start playing. Early on the game will put you in simple situations that require you to use one or more of the game’s mechanics. The game will then display a short description or demonstration of the mechanic you need to use to get over the current hurdle. The frequency with which the game shows these descriptions decreases over time. The iPhone games Spore Origins and Rolando contain good examples of this type of tutorial.
Both types of tutorials have their advantages and drawbacks. Mini-game tutorials are very focused. They allow users to learn everything at once. They keep out of the player’s way as they play the game. But what mini-game tutorials lack is context. In a game, context is less important because the world is rigidly defined. But in real-world systems, context is key to good user experience design. In-game tutorials are all about context, but they interrupt the flow of play. This is less of an issue in a game than it is in real-world systems. In a game, the frequency and temporal location of tutorial elements can be highly controlled. They appear when players expect them to appear, when beginning a game. The contextuality required to make these work in the real world means that they could interrupt important tasks and cause frustration.
Casual Game Design has several good articles on game tutorials if you are interested.
The Way Forward
I strongly believe that play is an integral part of the future of user experience design, and I am looking forward to making that future happen. To do that, I’m going to take the words of Mary Poppins to heart:
For every job that must be done, there is an element of fun. You find the fun and SNAP! The job’s a game!
I dub this the Mary Poppins Principle, and I challenge you to use it to find the fun in the jobs that your users must do. But for now, go have an ice cream cone. You deserve a treat for reading this whole thing.