The app has become synonymous with mobile computing. The phrase “there’s an app for that” wasn't coined for no reason. We all have our favourite apps, and everyone’s phone is customized differently. The app ecosystem has been integral to the success of the smartphone revolution. However, the app model has its critics.
Paul Adams, writing in Intercom, thinks the app will become a thing of the past, at least for the majority of use cases.
The experience of our primary mobile screen being a bank of app icons that lead to independent destinations is dying… The idea of an app as an independent destination is becoming less important, and the idea of an app as a publishing tool, with related notifications that contain content and actions, is becoming more important.
Adams argues that the shift from an app-centric model to a notification-centric (or card-centric) model will drastically change how we design apps and potentially the mobile operating systems. He envisions a world where instead of opening multiple apps to retrieve content, apps will push that content to a central experience - thereby providing us with a hierarchically organized stream of cards we can interact with.
The card interface was pioneered by Palm’s WebOS platform, but the design practices have quickly made their way into Android with Google Now and its new notification system, iOS with its task switching system, and Windows Phone’s tile interface. Adams feels that software developers are going to be developing software that integrates with card interfaces, which breaks “down [apps] into atomic units so that it can work agnostic of the screen size or technology platform.” He uses Facebook as an example, for it is not only an app but an entire ecosystem of aggregated content. He identifies a fast growing trend as “notifications as signposts, to containers (cards that include content), and actions on that content.” For example, Android and iOS users can answer messages and interact with other app content without ever having to open the app itself. In this regard, notifications become the entire experience.
However, this isn’t completely true because many apps (such as iOS’ Messages or Google Hangouts) only allow limited interaction from the notification pane. Users are not able to attaches images unless they open the full application. Admittedly, Android Lollipop will take this another step further as it will break down Chrome tabs and other atomic units into individual cards. Adams explains this as “apps as services.”
Breaking things right down into the individual atomic unit, including the content and actions. The atomic unit separate from the container of the app itself, so that it can show up anywhere, on any device. The atomic units are then reassembled based on context. Aggregated in a centralized stream. Or pushed to you on your watch.
The obvious problem with this model is that it requires a considerable amount of predictive computing to organize this seemingly messy interface into something that’s relevant to the user. The margin for error will almost certainly shrink. Adam does note that apps in their current form will still have relevance for deeper experiences and for content creation, but I would take it a step further. As a trained librarian, and one who works in the information science profession in higher education, I’m not sold on the idea of a card interface because it doesn’t reflect how individuals think. People do not organize their surroundings in cards. And anyone who’s ever used a rolodex or card catalogue will tell you that it’s organized very rigidly. Compartmentalization is an important aspect of our psyche. Our world has predictable organizational systems that make retrieving relevant information easy. A card interface that organizes our information for us using algorithms requires a lot of trust from the user.
From a graphic interface perspective, where is the user's the point of reference? App home screens - and particularly a home button - allow the user to get back to “home” from anywhere. A similar concept is the “desktop” of our home computers. Points of reference are very important because they form the foundation of the user’s conceptual understanding of the OS. Historically, mobile OS’ that don’t have obvious single points of reference require a considerable learning curve. Blackberry 10 is a perfect example since it relies completely on gestures and swipes. For this reason, a separate and modified Google Now/Siri interface makes more sense. Card notifications would reside in their own home screen or could be activated by pressing and holding a button.
At the end of his article, Adams asks whether card-based systems will take place at the app, notification, or OS level, whether there will be one or more streams, and who the owner of such systems will be. From a usability standpoint, I can’t see how a card based OS would make sense. My suspicion is that card notification systems will be built into the OS’ themselves, and each platform will have its own interpretation. I’m sure Samsung, HTC (which has already dabbled in this area with Blink Feed), and other manufacturers will release competing or customized notification systems for Android. Ownership will take place at the OS level, but third party developers might release their own on the app stores.
Another question I have about card notifications is the opt-in factor. As we’ve seen with Android Wear watches, the endless stream of card notifications can become unwieldy. Users can spend way too much time swiping away notifications because the platform isn’t very good at prioritizing what one wants to see. At least in the early stages of card notification systems, users will have to have a great deal of control over what they see, otherwise they will become overwhelmed and disable those features.
Adams’ vision of what card interfaces/notifications could be is quite intriguing. The idea that future notification systems would have embedded “child” cards within parent cards is quite possible. Twitter’s recent API “Fabric” is well timed, as it allows developers to utilize the Native Tweet and Tweet Composer tool - allowing for easier integration of Tweets into other types of content, on multiple platforms. Combine this parent/child relationship with card notifications that scroll horizontally and vertically, and the notification system becomes much more immersive. Users won’t always be prompted to open an app, and this is a good thing. Whether innovations like Apple’s Extensibility represent a move towards this vision is another question.
Adams also notes that apps aren’t going to disappear completely anytime soon because "opening an app is still necessary and great for many contexts, especially composition of new content and dedicated deep workflows, and maybe changing preferences.” He says notifications will be the place for finding new and interesting content, but again I disagree. Each app has a different “culture” if you will, and sometimes silos are good. I don’t always want to see what’s on Facebook, and often I’d rather just see Twitter or Google+ because the culture of those social networks and the type of content I get is different.
I also don’t think Adams is correct that apps are inefficient and that notifications will improve app discoverability. iOS’ Spotlight, Android’s built-in Google Now search, and voice commands make entering apps a breeze. I don’t even need to remember what folder I put them in. It’s true that the growing catalogue of apps on each platform makes discoverability difficult, but a barrage of notifications would only create a new problem. In fact, I would go so far to say that a notification-style mobile OS would replicate the current problems users experience with Email. Google has done some wonderful innovation with its email filtering system and its recent Inbox app for reaching “inbox zero.” But, a notification-centric OS is like creating an inbox OS. When users get tired of promotional email, they unsubscribe. I suspect when they get tired of app notifications they’ll uninstall apps. However, embedded apps might make this impossible. Users could see promoted tweets, ads, social media comments, etc and never be able to get rid of them. Adams’ vision leaves little room for error, and a more balanced approach would better suit users.
Emerging platforms - specifically wearables - are more inclined to make use of card notification systems because screen size constraints make a traditional multitouch interface impossible. I would argue that there will be design trends that make notifications more uniform across phones, tablets, and wearables, but the use cases with differ greatly depending on the device. Devices that have the ability to display lots of information (such as the iPad) will be more conducive to apps and large blocks of written content. This goes for phablets as well. While I think Adams is correct in saying that card-style notifications will play a large role in expanding the functionality of current notification systems, I don’t see the app disappearing. Card interfaces are a nice idea, but they don’t have users’ needs at heart.
Adams' full article can be viewed here: "The End of Apps as We Know Them"