Designing the Future: Context-Based Design

In his article “Apple Inc: A Pre-Mortem”, Dan M. touched upon the idea of Context-Based Design and how its implementation largely flopped in Apple’s current release of the iOS.

As an experiment, I believe that it’s constructive to analyze failures in design, and present a new way forward. So I opened a new document in Sketch to attempt to explore this concept. This exercise opened many doors and generated important design questions, which I seek to answer in this article.

What is Context-Based Design?

At its core, it is just an exponentiation of User-Centered Design and a progression on Beyer & Holtzblatt’s Contextual Design. It forces the designer to think thoroughly about the context in which the user will actually be using the product, such as the time of day, before or after which activities, and what the product itself can learn about the user, in order to enrich the experience.

It is the future. In a December 2016 article, Ted Livingston, founder of Kik, is right to say “chat is the new browser and bots are the new websites”. Bots are a great example of this new Context-Based Design because they provide a natural way to interface with services and connecting them more fully. I challenge Ted’s hypothesis by taking it one step further to say that the new browser isn’t a browser at all, but a full and complete integration of services (whether they are bots or otherwise) directly with the Operating System.

How do we implement it?

As designers and developers, our next move is to take a step back and observe the ecosystem and the tools available to us. I believe that today, the tools are all there in services like Anand Sharma’s Gyroscope and Apple’s Siri. We collect large amounts of data about ourselves, but it’s time to use that data to allow our most personal devices to know us uniquely and to tailor the experience to the individual. This must be represented by the entire interface changing based on how we are using and interacting with our devices. Below is an example of some context on my iPhone, which is not being leveraged.


Here you can see a heatmap representing the applications that I use most. In the current implementation of iOS, the springboard is a static screen. However, in a context-based iOS, the most used or most useful applications would be presented front and center. For example, “Pillow” could be hidden until night when it is time to use it.

The second point I’d like to make about the implementation, is that the functionality of apps is hidden behind the single tap on the icon. Why do we have to tap at all? In Apple’s most personal device ever, the Watch, they introduced something called “Glances”, which allowed the user to quickly view information and interact with apps. Rather than compartmentalize, integrate.

In a context-based iOS, apps would be displayed as “Glances”, which you can force press to access the whole app. Below, you can see a wireframe example of how this all comes together. At the top of the screen we have a “Context Bar” which displays your location, the time and date, and other contextual metadata. Instead of a static grid of apps, the UI now hosts a fluid array of “Glances”, which allow some minor, but useful interaction with apps without actually launching them.


The next piece of this puzzle is the learning process. This is what will enable iOS to adapt and give its user the most personal possible experience. Two types of learning take place. The first is “active” learning, which you can see on the bottom left. This takes place by the Siri system asking you questions in a non-invasive way. Additionally, “passive” learning will take place by the Siri system analyzing your usage patterns and any information which you allow it to use. It is important that Siri collects this data, because as a result, it will have more context about you, be able to better integrate with the OS and be more useful to you.

What about Notifications?

Finally, notifications are the next challenge. Have you ever felt that throughout the day, you are chasing down the little red dots on each app on your phone, trying to view and clear out these notifications, and the process is kind of annoying? The notification center provides zero respite, as it is a jumbled mess and it is difficult to sift through for the important information. This led me to conclude that there should be no notification center at all in a context-based iOS, but instead notifications themselves should be treated as another context. An additional piece of data should be added to each notification: a rating. Developers should rate their notifications by importance. Notifications which are assessed by the OS to not be useful to the user will have their rating downgraded so that they don’t bother the user. And the actual content contained within a notification should be shown on the home screen as a Glance.

So what am I hypothesizing?

In 2004, Geraldine Gay and Helene Hembrooke published “Activity-Centered Design: An Ecological Approach to Designing Smart Tools and Usable Systems” with incredible foresight about Context-Based Design. The introduction states that design at the time “at best assumed and at worst ignored the needs and preferences of end users”. This is entirely correct and I believe still characterizes the products which we use today, though to a lesser extent. In their own words, “another shift has begun now — to a context-based design…” And I believe that this shift has truly just begun. We are working on step one — to collect data about ourselves with tech-enabled watches and activity trackers. But it’s finally time to push that shift and begin using that data to design our products, with context.