Case Study: Go Outside!
From Meaningless Web Browsing to Reconnecting with Nature
Team: Aayush, Allan, Baker, Steve, Winston
In this 10 week course, we focused on encouraging folks to spend more time outdoors during the quarantine, in contrast to sitting indoors at a computer or scrolling on a phone all day.
Through a number of studies, experiments, and design work and usability testing, we’ve formulated a mobile app (and potential companion Chrome extension) to help people convert the breaks that we all spend online (whether browsing YouTube / Instagram / Reddit, etc) to ones where we get to spend more time outdoors, promoting both our physical and mental health [1].
Our mobile app targets folks who work from home and have not formed habits of consciously spending time outdoors. In our diary study, where we asked people how much they desired to go outside vs. recording how much they actually went outside, we found that reasons and goals for going outside were wide and varied (things ranging from “I want to get sunshine and see something green”, to “I want to take a walk to refocus on the next project at work”), we found that the excuses for going outside were relatively narrow: folks did not feel they had the time to do it, and many hadn’t built up the habit even if they felt they had time to (the inertia of other habits consumed the time instead). The other big excuse was that it felt like a waste of time, since they couldn’t come up with any interesting activities to do outdoors while social distancing.
Through our user studies, we found that an optimal time to nudge users towards going outside is when they are taking a break anyways. This helps to counteract the idea that “I am too busy to go outside”. When we determine that the user is taking a break, we will give the user a choice between two tasks that they may complete outdoors as inspiration. This helps to counteract the idea that “there’s nothing interesting to do outside”.
In the remainder of this case study, I will share some design challenges we encountered and solutions we’ve arrived at.
Challenge 1: Knowing when to go outside is hard
In our intervention study brainstorm, we eventually settled on an intervention meant to tackle the problem from a productivity lens: what if we created a “panic button” that folks could press when they felt unproductive, and receive a suggestion for something to do outside to clear their minds? We felt this would address the issue of having “no time” (if you’re feeling anxious and unproductive, a break can really help) and also the problem of having “no task” to look forward to outside. For the actual study, our Lo-fi prototype solution was to simply send folks a task to do each morning and a Google Form to upload a video or picture and short description of the task to share with us and (optionally) with others who had completed the task.
In our intervention study, we found our first big problem: very few people completed our tasks! Granted, our tasks were a bit difficult to do (one of them, record a “finger parkour” video, more or less requires a 2nd camera person to complete), but we found that even after repeatedly prompting our participants that many were hesitant to complete the task. On the plus side, folks who had completed the task said they had a lot of fun, and the pictures and videos they submitted seemed to match their comments.
Apart from the tasks being relatively difficult, things that may take about 30 minute to do, we realized that people weren’t sure when to do their outside tasks. Nobody had created a time block in their schedule to go outside, and asking someone for a daily 30 minute time block from their day was not so appealing.
Our design: reframe the activity
Instead of asking people to create time to go outside — something all our participants agreed was important on some level but found difficult to do — we realized it’s better to re-frame it as a healthier & superior alternative to scrolling through social media. This also has the benefit of targeting those “unproductive” moments throughout the workday that our participants knew they spent time on and wanted to replace anyways.
Challenge 2: how should tasks be chosen?
Given the low completion rate from our intervention study, we began to design around the tasks that we suggest to our users. How difficult or novel should the tasks be? How many tasks should we suggest at once? How should they be categorized? Should we give people a short questionnaire in order to personalize their task suggestions?
Our design: Decrease the importance of tasks & context is critical
While this is still a rich design area that is worthy of further study, we learned several things from our discussions and usability test.
First, we realized that the task is not the most important thing — instead, it’s to promote health and wellbeing through getting exposure to fresh air, sunshine, and vegetation. After interviewing our users after the intervention study, we found that the task simply didn’t appeal to some, while others already had outdoor routines they wanted to accomplish (e.g. running), and our task didn’t complement it in any way. So, we decided to offer each user two small “tasks” to choose from: less as an assignment and more as an inspiration for something to do that you could modify as much as you wanted. We chose 2 in order to give a choice, but also to prevent decision paralysis.
From our usability test, we realized we hadn’t taken into account weather conditions in our task suggestions! The tasks we suggest in a snowy, rainy, or scorching hot climate should be tailored to match the environment. This can be achieved by having the mobile app ask for the user’s location data, or notifying them that it will be approximated through their IP address, and is part of our next steps to improving our design.
Challenge 3: When to prompt users?
Given our learnings from the intervention study, we decided that finding the right time to prompt users was critical to habit formation and the success of our app. In our original “panic button” idea, we wanted to rely on users’ own mindfulness to know when they were feeling restless or anxious, etc, and press the button to get a good, refreshing outdoor break. But this proved to not work very well. On the other hand, asking folks to schedule times to go outside competes with all the other activities on their schedule, while the far-fetched idea of creating an AI to determine whether or not you’re being productive or not was just that: far fetched.
Our design: prompt users right before they take a distracting break
Our solution is to allow users to define a set of apps and websites that they frequent in order to take breaks, that they would like instead to replace with spending more time outside. This addresses the issue of relying on the users’ own awareness of their mental state to determine when to take an outdoor break, while circumventing the challenges of training a new AI method.
Challenge 4: notifications are annoying
Through our continued user interviews (and extensive personal experience!) notifications often result in a new habit of “ignoring the notification” rather than taking the action.
Our design: awareness of impact helps you gain perspective
From our usability test, we also sought to gauge how people felt about receiving the notification. We found that people were often not ready to immediately act on the prompt due to their context: in bed, on the toilet, or during a Zoom meeting, etc. We realized it will be important to not just tell people to choose a task and do it, but to provide awareness of the impact of taking a distraction break. In our future work, we may include as a part of our notification prompt “You spend an average of 1 hour and 3 minutes on Reddit each day”, helping people recognize the amount of time they are actually committing to the distraction break, and directly compare that to the tasks they are being suggested to do outside. We feel this will be a helpful way to gain perspective on how you are using your break time, so you can make more informed decisions about how you allocate your time. Additionally, we can use the context of the distracting app chosen to customize our task suggestion. For example, when opening Instagram, it could suggest users to take a picture that makes them feel peaceful, (vs anxious, jealous) etc.
Now, onto our top 5 usability challenges we found from our usability study!
1. Unclear Language
- We should name our “activity” prompts “distraction” prompts since activities are separate from prompts
- “Motivating Phrase…” had no clear meaning: we need some help text or even an example notification image to show the user what it does.
2. Form Inputs
- It’s preferable to use actual form labels to placeholders, since it’s easy to forget what each field represents
3. Onboarding
- Our app has a confusing first impression. The first screen of our app (the “View Prompt” screen) gave no indication or impression about what the app could do. This can be fixed with a short onboarding carousel with some cartoon UI graphics and text
4. Advanced Settings for Prompt Creation
- “Advanced” button is ambiguously named, and the advanced options page itself lacks a “Save” button (d’oh!)
- Unclear that it leads to repeat scheduled distraction prompts: would be better to place it all on one screen
5. Context-aware Tasks
- Tasks are not always feasible or attractive in the user’s current context, whether due to weather or their physical location (in bed, on toilet, in the middle of a Zoom meeting)
- Distraction prompts were confusing; people didn’t understand how the interruption would occur. Again, an example notification image would help clarify this!
[1] https://www.health.harvard.edu/newsletter_article/a-prescription-for-better-health-go-alfresco is one of many, many sources. It’s experientially true for most as well :)
