Wynd: A Portfolio Case Study
Creating consistency for an inconsistent personality. Wynd attempts to bridge the gap between winding down for bed and a college student’s typically sporadic lifestyle.
Sleep
Sleep is something that occurs to everyone on a daily basis. Especially during the pandemic, I’ve personally experienced (along with some select friends) “interesting” sleep cycles. For college students that don’t necessarily need to follow a strict sleeping schedule, we have the freedom to go to sleep at whatever time we please. Whether it’s falling asleep at 11 pm or 4 am, it seems to not matter when we only really need to get up in time for lecture the next day at 2:30 pm! While beneficial to sleep early, it’s simply not a priority for college students. So the big question is, how can we make sleep a priority?
Background and Research
Our team gathered 9 participants in a baseline diary study to better understand possible pain points that college students may have regarding sleep. In parallel, we also conducted some secondary research to brush up on our knowledge of the sleep field. Some points I found particularly interesting from the research:
- Sleep quality directly affects daily mood and productivity
- Existing tech solutions use static strategies (monitoring sleep and blue light reduction)
- Sleep is universal, but sleep quality metrics and habits are unique
- Mindful sleep prioritization leads to better sleep
Persona and Journey Mapping
Once the baseline study finished, we then synthesized our findings to create a typical user persona: Jeremy Rest. Journey mapping a typical day in the life of Jeremy Rest allowed us to identify opportunities for behavior change.
From journey mapping, a major finding that we found was breaking down the moments in a user’s life that affects their sleep. Particularly moments before bed, moments during sleep, moments after waking up, and moments throughout the day that can affect sleep later on. With that, we decided to hone in on moments before bed: particularly one’s evening routine.
Intervention Study
Our study consisted of two interventions to try and establish a nighttime routine for 11 participants.
- Set reminders on your phone as context prompts to remind participants about their goal of forming a routine. The intention is to disrupt the cycle of procrastination and overwork that usually happens during these times. Example set of reminders: 7 pm (5 hrs before bed, have dinner), 9 pm ( 3 hrs before bed, go shower), 11 pm (1 hr before bed, go wind-down)
- Pick a wind-down activity to conduct every night. This can include things like reading, playing a game, listening to music, or anything that relaxes your mind. We also had participants anchor this activity to something that is already a part of their nightly routine, such as brushing their teeth. By using an action prompt and anchoring, we hoped that participants would be more likely to adopt the wind-down activity as a habit.
Our main takeaway from intervention study feedback was that persistent reminders are annoying. Context prompts can be helpful, but not if it means stressing out the user with notifications.
Ideation
As a team, we came up with this laundry list of ideas. A fan favorite was a device that wakes you up with breakfast smells.
From these ideas, we eventually settled on an Apple Watch app that keeps you aware of your heart rate during evenings.
Prototype
From the idea, we were then able to create storyboards and wireflows of critical task flows of the app. Combining the task flows, we settled on Wynd. Wynd is a watch app that helps you wind down at night, keeping you on track to sleep at a reasonable time.
Features of Wynd:
- Onboarding introductory survey
- Checking in on your progress with maintaining wind down habits
- Nudges to ease you into bedtime through wind down activities
- Overactive alerts for when your heart rate exceeds a certain level
Usability Testing
By performing a usability test of our clickable prototype, we received very informative feedback. Of particular note was the language, UI, and navigation aspects of the prototype.
I’ll use our onboarding survey to highlight how the prototype failed in the aspects mentioned since it managed to miss the mark on all three aspects. In terms of language, categorical answers seemed ambiguous to new users. For example, as the designers of the app, we know fully well the distinctions between “Overactive”, “Wynd down”, and “Bedtime”. However, as a new user they have no idea what that means.
For caffeine sensitivity, from a UI and language standpoint, users weren’t sure what a 1 or 5 meant for sensitivity and which value means more sensitive versus less sensitive. Two different users could easily assume the opposite metrics.
Adding to that, at 13 questions, the onboarding survey was too long. Thus, a conclusion we made about the onboarding survey was that it needs major overhaul. It is the first thing that new users will see when using the app and these first impressions can really make or break the product.
Another example of UI and language confusion is in the habit tracking. Users weren’t sure what “Recalibrate” meant and some assumed that the light colored boxes meant completion when it was actually the opposite. They had to infer from the 71% or 79% that the purple boxes meant completion.
Below is a list of top problems we found from the usability test:
- Onboarding surveys should be short and sweet
- Onboarding process should educate user of how the app and its notifications work
- Button hierarchy, make a distinction between primary (filled) and secondary (outlined) buttons
- Navigation of the app should be seamless — have a “home page”
- Be clear with language: leave no room for ambiguity. User shouldn’t be playing a guessing game.
- Color consistency and contrast
Conclusion and Takeaways
This quarter has been a whirlwind of studies, journey mapping, system mapping, and collaborating on Mural and Figma. I learned that each step of the journey served a purpose and to not take even the smallest causal loops or quick and dirty sketchy screens for granted. You never know when it can yield some drastic, direction-changing insights.
If given more time, we would definitely have liked to make fixes to our prototype based on feedback from our usability tests. In addition, it would’ve been worth evaluating Wynd from an ethical standpoint and then making design fixes based on that evaluation. Instead of language or UI, it could be based on aspects such as inclusive design, universal design, and privacy.
Last, but not least, big thanks to the teaching team for sticking with us through this whole process and being so supportive!
