ScreenTime is Just the Beginning: An API for Digital Wellbeing

In the effort to build more humane technology, Apple’s recent inclusion of features to combat smartphone addiction in their mobile operating systems is a major step forward. These features, launched with iOS 12, will provide insights into phone habits, enable fine-grained controls for notifications, and will even allow individuals to self-limit device use.

For years, third-party companies like Freedom have been creating products that give users insight into, and control over their technology use. With Apple developing their own tools, the major players in the smartphone space have legitimized the cause of building for digital wellbeing. With features like iOS’ Screen Time, digital wellbeing has now gone mainstream.

And unfortunately, just as digital wellbeing has gone mainstream, Apple has shut down a number of major digital wellbeing apps – including Freedom.  While Freedom will eventually be back on the App Store, wellbeing apps will no longer contain powerful app or email blocking features.  ScreenTime, well intentioned as it is, will be the only game in town.

What problem does Screen Time solve?

When analyzing the new features released by Apple, we see they are built primarily around the problem case of addiction. And while smartphone addiction is a problem, it is not the only problem. Smartphone users want to be more productive and focused, be more calm and present, study and learn more effectively, and have control over their technology – among many other things. Fighting addiction is a positive first step, but it is only a first step.

look it doesn’t need to fit in your pocket if you never put it down

— Alexandra Petri (@petridishes) September 12, 2018

We face of future of ubiquitous devices and relentless connectivity. As we say in the Digital Wellness community, it is impossible for a single company to completely solve such a complex socio-technical issue as digital distraction.  For this reason, it is imperative that that Apple open up API’s so that third-party developers can build innovative tools for digital wellbeing that complement the offerings in the operating systems. By creating an API for Digital Wellbeing, Apple would signal a true commitment to the cause.

An API for a complex problem

Why should Apple create an API for Digital Wellbeing? Consider the scale of the problem. If you’ve got a smartphone in your pocket, you’ve likely struggled with managing your use. Our current smartphone era reminds of when cars didn’t have seatbelts and airbags – they are new technology with few safeguards. But unlike cars, which are the model of standardization, our phones are deeply personal, with varied use cases, with differing social and cultural modes of use. It is a disservice to the billions of users for Apple to roll out a few features and consider the problem of distraction solved (and then kick other apps in the space off the App Store).

What would an API for Digital Wellbeing look like? At Freedom, we’ve spent years studying the problem, and designing products that make technology less distracting. Leveraging ethnography, user studies, and analyzing big data, we’ve developed insight into the signals device makers can harness, and the controls they could enable in an API. This essay will provide a nontechnical explanation of the features that would go into an API for Digital Wellbeing.

How does Apple build it?

First, a spoiler alert: To effectively build an API for Digital Wellbeing, it’s not enough to make a set of on/off switches for apps. Maintaining a positive relationship with technology is an “embodied” experience, involving technology, social and emotional wellbeing, physical state, location, stress, and a host of other elements. Our devices track and make available much of this data, and it can be combined with other third-party sources like activity trackers (Fitbit, Garmin) and biosensors (Spire, Withings Aura). To build innovative tools for digital wellbeing, it is important that we’re able to leverage multiple sources of data.

Get insight, and then take action

There are two key concepts to an API for Digital Wellbeing: Insight and Action.

• Insight are the signals about our activity levels, physical state, our habits and routines, our technology use, that paint a picture of our state when using technology.
• Actions are the granular controls of the technology enabled by the API. These controls can include the ability to turn off apps and messaging, to silence notifications, and so on.

Being able to combine insight and action in an API for Digital Wellbeing would allow developers to create powerful, enabling solutions for the billions of device users out there.

I’ll map out some of the major concepts within Insight and Action, to illustrate some of the ways application developers could leverage an API for Digital Wellbeing. Before I do that, it is important to note that any kind of design for digital wellbeing should be privacy conscious. Our activity data is amongst the most private and personal data we generate, and should be afforded privacy protections. These API’s should be opt-in only, with appropriate informed consent, a right to withdrawal of information, and appropriate privacy safeguards.

Insight: Technology Use

A primary concern when designing for Digital Wellbeing is understanding technology use. To build solutions that enhance Digital Wellbeing, we should understand an individual’s areas of use and overuse of technology. Apps like Moment, and Apple’s own Screen Time provide insight into certain types of technology use. An API could provide insight into the apps and websites that account for time spent, insight into time on device, device pickups and unlocks, and other measures of device use.

Insight: Place

Where we use our devices is very informative – consider how your device use changes based on place. You may use your phone differently at home and work. You may go to the same coffeeshop to focus and write. And our technology behaviors will change throughout the day, as we move from place to place (e.g. focusing on communicating and scheduling in the mornings at work, browsing and exploring in the evenings at home). For these reasons, being able to access data on place is a critical component when designing for digital wellbeing.

Insight: Routine

Our device behaviors are highly routinized. Think about the laptop you use during the workday – there is a strong routine and pattern to that device’s use. And to make things more granular, on those devices, there are certain apps and websites you may visit with a routine. For example, news websites in the morning, or workout apps after you finish exercising. The digital trails left behind by our routines are highly useful when designing for digital wellbeing. Imagine apps that provide insight into routine, help understand deviations from routine, and can guide you back into productive habits.

Insight: Physical State

Our physical state – how we feel, how much sleep and exercise we get, our levels of hunger or tiredness – is affected by, and affects our device use. Think about how your willpower to focus differs between when you are energized versus when you are tired or hungry. Or think about how device overuse can affect your physical state – keeping you up at night, etc. An API for Digital Wellbeing would incorporate a range of physical measures so they could be correlated with device use. In doing so, better controls and recommendations could be developed to ensure that devices are contributing to improving our health and wellness.

Insight: Emotional State

How does device use make you feel? What apps make you happy, and what websites contribute to feelings of unease or sadness? While our devices may not be great at measuring our emotional state, there’s an important connection to be made between device use and emotional state that should be an important component of an API for Digital Wellness. Understanding the apps that improve our emotional state can lead to more fulfilling time on devices.

Insight: Activity

Our devices should help us get things done – plain and simple. For this reason, measures of activity – words typed, articles read, achievements – can inform our understanding of our positive and negative device utilization. Understanding when we are productive, and how productive we are is important. While “activity metrics” are generally simplistic (e.g., I may be typing a lot, but what I type may not be particularly high quality), they capture relevant aspects of our interaction with technology.

Action: Controls

Now, let’s switch to the kind of actions we’d take based on the insights generated from our mobile devices. Apple’s ScreenTime enables you to translate insight (app overuse) into action by blocking the app. At this early stage, most of the actions you’d take are some form of control. These controls could involve blocking the app permanently, temporarily, or on a scheduled basis. Blocking of an application is probably simplest, grossest notion of a control.

As an API for Digital Wellbeing became more robust, you’d want to selectively block features of applications (e.g. turn off notifications, block the browsing of certain websites, block certain messages). This would require developers building hooks to the API within applications – but this is not a far-fetched notion, as we’ve already seen Facebook experimenting with notification controls within their application.

At Freedom, we developed network filtering and other technologies to turn off certain parts of applications.  These filters have proven to be very popular – but the logical next step is to work with developers to enable API-based controls within applications.  Imagine being able to turn off Facebook Messenger for a few hours a day from a simple interface?

Action: Strength

Now we get to one of the more interesting, and tricky components of an API for Digital Wellbeing. If you are going to give people control over their devices, that control will need to have a variable level of strength. Consider the case of Apple’s ScreenTime. In ScreenTime, if you go over the time allotted to an application, you are blocked from launching that application. But, you can also choose to circumvent the block. While this level of strength will likely suit many people, some will want their blocks to be “stronger” – impossible to circumvent.

Strength is an important concept because many tools for digital wellbeing are configured in one mental state, and experienced in another. That is, when you decide you want to be blocked from Instagram after an hour, you’re exhibiting high willpower. When the blocks actually kick in, however, you may not have the same amount of willpower – and without an appropriate strength for the block, you may just decide to circumvent.

We’ve learned through years of research that people need better options for self-management – which is why we include Locked Mode in Freedom.  Locked Mode allows the user to decide on the strength of their blocks – including making them impossible to circumvent.  By missing this use case, Apple has significantly undermined ScreeTime.

The notion of strength is particularly challenging – one can imagine being shouted down at a product meeting when trying to make the case that, yes, people do really want to be locked out of their device for the day. But strength is essential in that it cements our intention in a way unique to technology – we can make a decision and technology can help us follow through.

Putting it all together

We can think of an API for Digital Wellbeing as similar to Apple’s HealthKit.  Apple becomes an honest broker for the insight data, properly anonymizing and safeguarding it, while third-party apps leverage it to perform actions allowed by the OS.  Using this API, an app could create a rule that allows you to turn off distractions every time you go to your favorite coffeeshop or coworking space.  Or an app could leverage physical and emotional signals, identifying times to turn off stress-causing apps.  The possibilities are endless – and very few are supported by the Apple’s basic ScreenTime implementation.

But would Apple build such an API?  It is clearly possible – Apple already has access to all of the activity data, and now has built the actions into the iOS 12 operating system.  There are no technical limitations at this point – it is simply an organizational choice.  And it is a choice that would signal a meaningful interest in solving the problems created by these pervasive devices.  It is our hope that ScreenTime is much more than marketing material, and that Apple has plans to expand these initiatives to truly help users.  Of course, kicking digital wellbeing apps off the App Store does not seem like a promising start – but we’re not going to give up on Apple yet.  The problems created by pervasive technology are some of the biggest technological challenges we’ll face, and by building an API and opening the doors to developers, Apple will signal a true commitment to solving the problem.

In the meantime, the team at Freedom will continue to innovate and lead in this space.  We’ll be putting out a series of articles about how ScreenTime can be improved, and elaborating more on the API for digital wellbeing.  It is our hope that this is only the beginning of a conversation – a very important conversation.