Michael Biskjaer, Peter Dalsgaard, Kim Halskov
Find a spartan room with a clear desk. Wear a pair of earplugs and over them noise-canceling, pink-noise-emitting headphones. Use a modified computer with no card for computer games and with the Ethernet port sealed to block Internet access. Now draw the curtains—and put on a blindfold. Then you have re-created what author Jonathan Franzen considered his distraction-free digital haven when he wrote his acclaimed novel The Corrections (2001).
Extreme as this example may seem, it shows the lengths to which some individuals will go in order to fight what we consider an emerging trend in HCI: a significant increase in digital distractions. That the Internet’s abundance of digital information—be it news, discussions, memes, music, videos—is alluring and may distract users is nothing new. However, adding to this type of distraction, a new type has emerged in which the user is rendered passive. This category is digital information as ubiquitous, instant notifications, reminders, alerts, and alarms that seemingly pop up out of the blue and demand the user’s attention during activities entirely unrelated to this information. In this scenario, we are no longer talking about Mark Weiser’s envisioned age of calm technology, in which ubiquitous computing gradually recedes into the background of our everyday life. Rather, we are talking about a hectic, chaotic space in which various digital technologies compete for the user’s attention and take a toll on the user’s ability to employ the technology for the very purpose it was originally brought into play. Indeed, the technology seems to drive the user, not the other way around.
While this first trend—the increase in unbidden digital distractions—is highly relevant to the HCI community, we have noticed another and possibly even more intriguing trend. This emerging second trend can be seen as a reaction to the former. It is characterized by software designed to decrease, delimit, or discard exactly such unbidden digital distractions by limiting, blocking, or even entirely removing the very opportunities for conveying digital information that the other types of software are designed to afford. Using this new type of purposefully constraining software resembles the act of hanging a Please Do Not Disturb sign on a hotel-room doorknob. We therefore call this type of software Do Not Disturb software (we offer a more detailed definition later). In this article, we explore these two opposing trends and offer a basic typology and a framework for understanding some of the salient aspects of how Do Not Disturb software is, rather ironically, designed to help ensure digital freedom through digital constraints.
We see this first trend as being characterized by at least two elements: device-based notifications and the Internet’s constant stream of social media notifications.
The first element is the growing number of smart devices seeking to grab the user’s attention to inform them about current system status, updates, and so on. Here, notifications are a prime example. Previously, digital devices came with no pre-installed notification software. Therefore, users wanting this type of information had to resort to add-on notification software such as Growl (released in 2004 by a development team led by Christopher Forsythe), which is installed as a preference pane added to the Mac OS X System Preferences, thereby allowing users to customize their desired notifications. Today operating systems such as Mac OS, iOS, Windows, and Android come with a built-in notification center to handle the increasing number of notifications from both manufacturer-installed and user-installed software, as well as notifications from the OS itself. This design is trickling down to other and more specific types of software; for example, browsers such as Chrome, which let individual websites and extensions notify users of updates. This trend is only likely to accelerate as Internet of Things (IoT) devices and wearables such as smart watches increase the number of notifications and the potential for notifications to reach users through interfaces. Although visibility of system status is considered one of the key principles of usability, a pressing question is how much information about device and system status users can (and indeed want to) process—and at what cost.
The second element in this first trend is the Internet’s social media and the many novel ways of creating, sharing, browsing, and checking information that it has enabled. Some users consider the constant availability of Facebook and Twitter an almost invasive force in their daily lives, and Baron  has suggested that users tend to regulate their accessibility via continuous partial attention, given the strong presence of these online social media. With an estimated average of approximately 500 million tweets per day, the potential for digital distraction on Twitter alone is huge. This intense online activity is mirrored in a continuous stream of social media notifications informing users about other users’ status updates, the latest cat memes, fun websites, breaking news, weather reports, and much more.
Given this increase in digital distractions both offline and online, it is hardly surprising that users of computing technology have resorted to various strategies for overcoming this force. Earlier we mentioned Jonathan Franzen’s extreme individual approach. Other, more social-oriented and likely more familiar strategies would include agreements among family members and friends stating that a person—say, while traveling or attending a meeting—cannot be reached or notified digitally about any event unless it is urgent. Other well-known examples are company policies requiring staff members to check their e-mail only at certain times during the day, and the quasi-movie-trailer in theaters instructing the audience to mute or turn off their devices during the show.
A pressing question is how much information about device and system status users can (and indeed want to) process—and at what cost.
What we focus on here is the rise of specific digital strategies. Do Not Disturb software is thus designed to handle digital distractions by constraining the very opportunities that other types of software are designed to enable for the user in terms of either letting them (intentionally) access or (unintentionally) be informed of digital information. Recent years have seen the emergence of many such Do Not Disturb applications to regulate—in other words, reduce, block, or even remove—the affordances of other applications in order to overcome the problem presented by the increase in digital distractions.
The best-known example of Do Not Disturb software may be Freedom (Figure 1), an Internet-blocking app with more than 500,000 users including award-winning writers Nora Ephron, Naomi Klein, Nick Hornby, Zadie Smith, and Dave Eggers. Launched in 2008, Freedom was developed by Fred Stutzman, who at the time was working on his doctoral dissertation in information science at the University of North Carolina at Chapel Hill. Stutzman has explained that he was inspired to build the app when the coffee shop he used to frequent for efficient writing time one day began to offer free Wi-Fi. Soon his focus on his dissertation abated due to the magnetism of the Internet, namely its opportunities for inspiration and diversion. To get back on track, Stutzman built Freedom as a means to transform his computer “from being a source of distraction to a device of work” . By a “brute-force hack,” Freedom in its standard configuration (freeware) blocks Internet access for a period of time set by the user. Once installed, the app sits in the menu bar. If the user needs to go online during a session, rebooting the computer overrides the app’s offline setting and restores full Internet access. Today, Freedom is no longer unrivaled—many related applications now promise to help users take action on distraction.
Inspired by the success of Freedom, we have conducted a preliminary study of various types of Do Not Disturb software tools and strategies. This has been carried out as part of an international research project, Creativity in Blended Interaction Spaces (CIBIS; http://www.CAVI.au.dk/CIBIS), focusing on how and to what extent digital technology can enable—and restrain—creative practices. In order to locate some of the most popular examples of Do Not Disturb software, we have confined our research to five areas: Apple’s App Store and iTunes, Google’s Play Store and Chrome Webshop, and leading online technology magazines. From this pool of examples, we have selected 10 for the reasons that they have reached a wide audience and show diversity in their feature sets. On this basis, we offer a basic typology and a preliminary framework for understanding key features of Do Not Disturb software tools and strategies.
Device level. Before we discuss the 10 software examples, it should be mentioned that the user’s choice of hardware can act as a strategy to accomplish the same goals as Do Not Disturb software. One purely hardware-based strategy for managing digital distractions is to use an obsolete device, e.g., one that cannot go online or does not support alerts or notifications (this is related to the HCI topic of non-use). This strategy renders Do Not Disturb software redundant. Other hardware-based options include sealing the Ethernet port or removing the PC’s wireless card, which was Jonathan Franzen’s strategy. Interestingly, recent years have seen a new type of deliberately constraining hardware such as Freewrite (previously called Hemingwrite), which was introduced in a Kickstarter campaign in 2014 (Figure 2). Freewrite resembles an alternative typewriter. It features an aluminum body with mechanical switches and a small e-ink screen. The device can store the written text documents both using the built-in storage and in the cloud. Similar to the app Freedom, though, Freewrite cannot go online for the purpose of browsing.
System level. This level concerns preinstalled general OS settings. In addition to the now common built-in, customizable OS notification centers, a familiar example is the iPhone’s Do Not Disturb mode in iOS. This mode enables users to silence calls, notifications, and alerts while the phone is locked, and to schedule a time when calls are allowed. Software on this system level bears a resemblance to individual strategies for managing digital distractions, such as muting all system sounds, deactivating Internet access, and switching off all alerts and notifications.
Application level. The application level involves the deliberate use of specific individual apps. Do Not Disturb software tools and strategies on this level include the system-specific app Freedom. Other examples are task-specific apps. By this we mean apps that do not block online access, but rather more subtly address the issue of distraction from complex (overly feature-rich) interfaces. A familiar example is the writing/reading Focus View in Microsoft Word, which aims to provide an interface with minimal distraction by presenting only a white sheet of paper on a customized background. Switching to another app window or simply pressing the Esc key aborts Focus View. Other apps offer a similar, highly reduced interface; for example, iA Writer, which, according to the developer’s website, aims to reduce form and content to their essence.
A more radical example is Flowstate, owing its name to Mihály Csíkszentmihályi’s concept of flow as a state of complete absorption in a given activity. What makes Flowstate stand out as an app is the fact that if the user stops writing for more than five seconds in a writing session (whose length the user decides), the written text will simply disappear from the page, thereby resetting the timer to begin a new session. Do Not Disturb software tools on this application level can thus be compared to individual strategies for managing digital distractions such as turning off an e-mail app or browser, uninstalling non-work-related apps, choosing apps with very few features, using advanced apps in simplified mode, and so on.
In-application level. Detailed in-app settings is the most advanced level. We have discerned two subgroups of such Do Not Disturb software tools. One is browser extensions such as Time’s Up! Facebook Time Limiter for Chrome, which notifies the user of how much time they’ve spent on Facebook and turns off access to the website after a user-defined period. The other subgroup is standalone apps with detailed in-app settings; for example, Anti-Social, which, as opposed to Freedom, blocks only social websites, and SelfControl, which lets the user block specific websites for up to 24 hours. What makes SelfControl radical is that it cannot be turned off once set—neither rebooting the PC nor deleting the app helps. The blocked sites stay off-limits until the SelfControl timer runs out.
A more advanced example is Rescue Time. This app runs in the background and tracks the time the user spends on various apps and websites, which they can set to be blocked. This data is converted into a report to give the user an accurate picture of their day. Do Not Disturb software tools on this advanced level are therefore related to individually tailored strategies for tackling digital distractions; for example, the user may permit alerts, notifications, and messages from handpicked people (favorites) only, or set up an out of office auto-reply message even when they are available. We sum up these examples of Do Not Disturb tools and strategies in the tentative, basic typology shown in Table 1.
Having perused the above 10 and many other examples of purposely constraining digital tools and strategies for taking action on digital distraction, we define Do Not Disturb software as: software that on the system, application, or in-application level is designed to deliberately constrain the user’s freedom of action by reducing the potential for digital distractions in order to support the user’s focus on the intended task. Although our examples differ in scope, they share some salient features. We have singled out three for the sake of comparison.
The first feature concerns what we call the level of constrainedness. As a concept we have explored in depth elsewhere , using purposefully constraining software means exerting a self-imposed constraint in order to gain a benefit. Accordingly, we define the level of constrainedness as: the applied software’s constraint intensity, which governs what the user can and cannot do while carrying out the intended task. The second key feature is the user’s ability to customize the software, so we denote this as the level of customizability. Our definition: the extent to which the user can adjust the software to fit their individual preferences and requirements for the purpose of carrying out the intended task. Finally, we consider user effort when using this type of software. Our definition: the effort required from the user to activate and employ the software for the purpose of carrying out the intended task. We briefly review our 10 examples of Do Not Disturb software based on these three key features and present our preliminary findings in Table 2. Since this is meant as an overview and proposal for more in-depth studies, we categorize our examples as low, medium, or high for each of the three features to display their interrelational diversity.
As for constrainedness, built-in OS notification centers and their mobile equivalent, iOS Do Not Disturb mode, and MS Word Focus View all feature low constraint intensity. The two former can be engaged or deactivated using a switch button (presenting two mutually exclusive choices or states), while simply pressing the Esc key or switching to another app window aborts Focus View. The writing app iA Writer may resemble MS Word Focus View, but since its intentionally minimal interface is constant, we place this is in the medium category of constrainedness alongside Freedom, which is also clearly constraining but can be deactivated through rebooting. The Time’s Up!, Anti-Social, and Rescue Time apps offer a significant level of constrainedness through user preferences, but still feature an easy exit since the apps can be turned off. A high level of constrainedness is found in Flowstate and SelfControl, since the former deletes written text if the user loses focus on writing for more than five seconds, and the latter allows for no deactivation at all.
Similarly, our 10 examples differ in terms of customizability: MS Word Focus View, iA Writer, and Flowstate are all task-specific apps focusing specifically on writing, which is mirrored in their low level of customizability. Apart from the iOS Do Not Disturb mode and the system-specific app Freedom, the medium group consists mainly of apps that allow the user to block either Internet access for certain periods of time or certain user-selected (often social) websites. This includes Time’s Up!, Anti-Social, and SelfControl. A high level of customizability is found only in the built-in OS notification centers and in Rescue Time, with the latter targeting quite demanding users by offering precise, user-defined blocking of websites, apps, and notifications, and detailed reports and weekly e-mail summaries of the user’s working habits.
Finally, Do Not Disturb software requires dissimilar amounts of user effort. Activating MS Word Focus View can happen through a keyboard shortcut. iOS Do Not Disturb mode is built in and requires but a tap on a switch button and a few very basic settings. Equally low levels of user effort are involved in Freedom, iA Writer, and Flowstate, as they work “out of the box.” Built-in OS notification centers can be used as is but are designed to be set up based on user preferences. This suggests a medium level of user effort. The same goes for Time’s Up!, Anti-Social, and SelfControl, which require the user to blacklist specific websites. An even higher level of user effort is found in Rescue Time, which comes into its own when the user meticulously defines their preferences among the many available options.
As shown in Table 2, our brief survey presents some of the dissimilarities between our 10 examples of Do Not Disturb software based on three (among many possible) features: constrainedness, customizability, and user effort. Although Time’s Up! and Anti-Social have the same profile, the software examples are all comparatively highly diverse. Interestingly, we note that—to the best of our knowledge—there is currently no Do Not Disturb application available that offers a high level of both constrainedness and customizability. This suggests the need for more theoretical work as well as empirical user studies to gain better insight into the potentials of using Do Not Disturb software, especially by focusing on discerning, defining, and developing key principles for designing this type of software. Here, recent studies in attention theory in psychology will be valuable; for example, the finding that task-unrelated distractors can indeed capture the user’s attention and disrupt task performance, but only in conditions of low cognitive load. If the user is sufficiently challenged in their activity, any task-irrelevant attentional capture is eliminated . Another important perspective for designers is the user’s individual rhythm of focus and distraction during the day; rhythms of attentional states are related to work context and time of day .
In closing, we mention that our observation of an increase in digital distractions is reflected quantitatively. A 2015 Global Web Index report (http://insight.globalwebindex.net/social) shows that globally, the average Internet user spends approximately 1.77 hours each day on social networking—an increase from 1.61 in 2012. It thus seems very plausible that both the relevance of studying and the need for designing Do Not Disturb software for taking action on distraction are only likely to grow.
This research has been funded by the Danish Innovation Foundation grant 1311-00001B, CIBIS.
2. Zax, D. Can we achieve ‘Freedom’ from our iPhones? MIT Technology Review (Dec. 8, 2011); https://www.technologyreview.com/s/426319/can-we-achieve-freedom-from-our-iphones/
4. Lavie, N. Attention, distraction, and cognitive control under load. Current Directions in Psychological Science 19, 3 (2010), 143–148; http://dx.doi.org/10.1177/0963721410370295
5. Mark, G., Iqbal, S.T., Czerwinski, M., and Johns, P. Bored Mondays and focused afternoons: The rhythm of attention and online activity in the workplace. Proc. of the SIGCHI Conference on Human Factors in Computing Systems. ACM, New York, 2014, 3025–3034; http://dx.doi.org/10.1145/2556288.2557204
Michael Mose Biskjaer is a postdoc at Aarhus University. Building on interaction design, philosophy, aesthetics, and creativity research, his work applies an interdisciplinary humanistic perspective on central theoretical themes in individual and collaborative creative processes in art, design, and innovation. He currently studies how analog and digital constraints may improve creative performance. firstname.lastname@example.org
Peter Dalsgaard is an associate professor of interaction design at Aarhus University. His work combines practice-based experimental interaction design with theoretical developments in understanding design processes and the interplay between the digital and the physical. His current work focuses on improving our understanding and use of digital tools in creative work processes. email@example.com
Kim Halskov is a professor of interaction design at Aarhus University, where in addition to being director of the Centre for Advanced Visualization and Interaction (www.CAVI.au.dk) he is also principle investigator of the Creativity in Blended Interaction Spaces project (www.CAVI.au.dk/CIBIS). His research areas include innovation processes, design processes, and experience design. firstname.lastname@example.org
©2016 ACM 1072-5520/16/11 $15.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2016 ACM, Inc.