How To Design a Mobile OS
This week we a saw a new version of iOS, one the world’s most successful mobile operating systems. When the original iPhone came out, the OS in some ways was light years ahead of other phones.
But we are only 5 years into the smartphone world and there many many things for us to learn still. Both Android and iOS are not the last word in mobile operating system design.
1. Get out of the users way.
At it’s core an operating system’s purpose to help users get something done. Whether its to make a phone call, connect to friends or get the latest news, how quickly and easily a user can accomplish a task is the number one goal.
A simple metric, taps to completion is a good rule of thumb.
For example on iOS, to make a new phone call.
- Tap 1 - Tap The Home Button
- Tap 2 - Tap The Phone App Icon
- Tap 3 - Tap The Keypad Button
- Tap 4 - Enter the number
- Tap 5 - Tap the Call Button
5 Taps to Completion
Compare this to older dumb phones
- Tap 1 - Tap the Green Call Button
- Tap 2 - Enter your number
2 Taps to Completion
Granted maybe some people are making less phone calls these days but I still use my phone as a phone all the time. What if mobile phones had a quick action button? One that the user could specify to always do the same thing, whether launch the phone dialer or take a picture or text someone?
2. Provide Joy
We live and sleep with our phones these days. Something that is so close to us needs to provide joy and delight in our lives. It’s not just a matter of getting things done, but getting things done with a smile.
Animation, sound and visuals are a great way to provide joy. When you first see the great animations bringing to life the experience of your phone, it provides a human touch to another wise cold and lifeless device. Humans move, breath, play music. We are alive with emotion and respond in kind.
A simple metric, smiles per interaction is a good rule of thumb.
That is to say, for every interaction (calling someone, changing apps, responding to messages), how many times does the user smile? Granted the smile most often only happens the first time, but I still smile every time I flick a business card with CardFlick.
This maybe a harder metric to measure or design for, it’s one of those you know it when you see it/feel it kind of things, but that’s what the emotional part of Human Computer Interaction is about.
It’s how our devices relate to and create experiences for us as humans that matter.
When a user first opens your app or your OS are they smiling? Does the home screen experience immediately radiate quality and joy?
3. Keep My Context
Phones are also unique in that they are almost always not the focus of your attention. We are walking with them, waiting for the bus with them, hopefully not but still happens driving with them. Keeping the user’s location and navigation options clear at all times is key for a great user experience.
A good metric is steps to home and back.
That is how many actions must a user do to get to a neutral state and then return back to the task at hand.
On iOS
- Tap 1 - Tap the home button
Gets us home. Awesome.
- Tap 2 - Click the app I was on if on home screen.
Once again awesome because most apps remember where you were.
Or
- Tap 2 - Double Tap Home Button to bring up running apps.
- Tap 3 - Click the app I was just in.
Really great when it comes to going home, but I think there are ways to improve returning to the last app used.
What if a single tap on the home button brought you home and a double tap returned you to the last app?
4. Be Consistent
There’s a reason most door handles are in the same place. It makes it very easy to understand how to open a door. Compare that to swinging doors that have push or pull written on them, how much human energy is wasted on figuring that out and bumping into glass?
The same thing applies for user interfaces. Consistency breeds usability. The same kind of actions in the same kinds of places. The same font and button styles used to indicate similar actions.
A good metric for this is number of styles per screen.
For example the wonderfully useful control center in iOS 7 has at least 4 different kinds of buttons.
- Round Toggle buttons for airplane mode, wifi, bluetooth etc.
- Rectangular Action buttons for AirDrop and AirPlay
- Square Toggle buttons for flashlight, camera, calculator etc
- Icon Action buttons for play/pause without borders.
5. Design for Multiple Users and Multiple Places
We use our phones on the subway in darkness, at the park in bright light, in our cars where we can’t see them. Some of us have older eyes with color vision problems, and some of us are sharp as a tack. The design of a mobile OS should encompass all these kinds of users.
One of the key ways to do this is provide contrast.
Elements that don’t look similar if they have different functions or have typographic contrast make usage much more clearer. Buttons that feel interactable whether with outlines or shadows or other visual focusing tricks really help. Font choices that make use of boldness, size and placement can lead visual heirarchy to make UIs more understandable.
A great way to test this out is the squint test. Look at your UI and squint your eyes, are the actionable elements still obvious?
6. Grow with me
As we get more and more comfortable with touch screen interactions, the general populace is able to understand do more advanced interactions. Where pinch and zoom was magical and new a few years ago, it’s now standard. Many apps before iOS 7 used a left swipe to go back to previous screens and now it’s built into the OS.
The lock screen on many mobile operating systems is just a bunch of stickers. Not actionable or engaging. And while a giant clock on the face is very useful, I think we can also think of other things that can be useful on the lock screen. Actionable notifications, last unread message or call, heck even showing the last good photo I took. Change over time, static is boring.
Both iOS and Android have amazing app stores where much future functionality can be added in. But I think we are a point now where the default experience has much further to go before we can consider it revolutionary.
