Mojo Vision: AR Contact Lens Notifications Case Study

My role: Solo UX Researcher

Timeline: ~4 months total

• Part 1: March and April 2022 - 5 weeks

• Part 2: May and June 2022 - 5 weeks

• Part 3: September and October 2022 - 6 weeks

Stakeholders: Mojo Vision Executives and Designers, future users of Mojo Lens

Methods Used: Eye-tracking, In-person Usability Testing, Semi-structure interviews

Context:

Mojo Vision (a start-up) was creating an Augmented Reality contact lens that would have been interacted with solely through eye-gestures. A critical feature for the Mojo Lens is to be able to receive notifications and respond to notifications. Before this research project was tasked to me, there was a design prototype that implemented a basic form of notifications. This prototype was constructed with no previous UX research and was just used to serve as a proof of concept. After a few people interacted with this prototype, it was evident that these notifications weren't the best or most effective. At this point, the design team was wondering: “how should we show notifications on the Mojo Lens?” and “what should the notification experience feel like with this lens?” I was tasked with conducting research to gain insights and improve the notification experience on the lens.

Overall Business Goals:

• Create a notifications feature that makes receiving notifications on the lens enjoyable, easy to interact with, and fits in with the rest of the lens UI

• Figure out details of how notifications should look and act on the interface in order to have a smooth experience

The business goals were slightly different for each part of the research study.


I decided to break this broad task into three parts, where the research in the previous parts informs the direction of the next part of the research. 

At a higher level, these three parts were:

• Part 1: Notification Reaction Test (March/April 2022 - 5 weeks)

• Part 2: Notification Animation and Positioning  (May 2022 - 5 weeks)

• Part 3: Start to Finish Notification Usability Testing (September 2022 - 6 weeks)

Part 1: Notification Reaction Test 

Introduction:

This was a behavioral level test to see how people react to a stimulus. Because of technology constraints and decisions, the Mojo Lens interface was a particular case; no other previous study regarding visual stimulus and eye-movement was directly applicable. In this case, a literature review was not helpful and I decided it would be worthwhile to conduct our own eye-tracking test with the Mojo Lens constraints.

Part 1 Primary Research Questions:

• What is the best size, brightness, and location to show notifications in a way that is the least distracting?

• Are single pixels or small dots still perceived?

Participants:

(Constraint: Can only test with employees due to confidentiality concerns)

10 participants of different age groups, intentional since eye movements and perceptions change with age.

Part 1 Methods:

• Eyetracking study

Participants were each shown 54 dots. These dots were different combinations of brightness levels, sizes, and locations that we were testing. Participants were watching a video with audio while these dots appeared in their vision. If participants noticed a dot, then they clicked a button that confirms that they perceived the dot. Participants were told to try to not look at these dots, or not let themselves be distracted by these dots, and fully pay attention to the video in front of them. In this study, I was tracking their eye movements to see how long participants looked at these dots when told to avoid looking at them. The assumption in this study was that with the ideal size and brightness, participants would be able to perceive the dot but not be distracted by the dot. 

Part 1 Results and Impact:

After looking at the eye tracking data and corresponding the duration of the eye movement and button clicks from the participant, I was able to make an argument for ruling out certain dot sizes and brightnesses. This research finding directly translated into a key metric of how big and bright the design team made our notification icons. The quantitative data resulting from eye gaze data also impacted the timing of some aspects of the interaction design. 

Most importantly, this research study also changed the way the design team and I conceptualize notifications. Before this study, we had several assumptions on what we thought people would perceive and how their experience might be. These assumptions did not seem to be the case as shown by the qualitative data I gathered. We changed our way of thinking about the concept of notifications on the Mojo Lens after this research study, and it helped me frame the way I would ask questions to participants in Part 2.

Part 2: Notification Animation and Positioning 

Introduction:

The actual notifications on the Mojo Lens was going to be in the form of icons rather than just dots. The goal part 2 was to understand people’s perceptions of, and interaction with, those icons.

Part 2 Primary Research Question:

• How should notification icons be conceptualized and animated to fit in with the rest of the lens UI?

Participants:

(Constraint: Can only test with employees due to confidentiality concerns)

12 participants who have not seen versions of these current notifications. 

Part 2 Methods:

• Semi-structured interviews, eyetracking, and usability testing study

This was a purely qualitative study. Because I wanted to understand how participants were perceiving these different concepts and animations, it didn’t make sense to conduct a more rigid A/B test. I gathered data in two ways for this study. The participants explained their experience and their feedback through a semi-structured interview format at the end of the session, and were instructed to share their feedback while interacting with the different concepts. I asked the participants questions during the session that let me gather insights on their actual perception of the icons. To provide a high level example, participants were asked what shape they saw and if they said “circle” when they were actually shown a square that would be marked down. 

Part 2 Results and Impact:

I was able to make design decisions from analyzing the data I gathered about what the participants actually perceived and their feedback about the experience. After analyzing the feedback and experiences of the participants, I saw the benefit of making a major UI design concept change. I presented these findings company-wide and advocated to change multiple big aspects of the design based on the findings of this study. After my presentation, the rest of the design team was on-board with prototyping this concept shift I was advocating for. The goal was to test this second prototype in the future to see if it does improve the user experience. We incorporated this new concept into the design project roadmap. 

Part 3: Start to Finish Notification Usability Testing 

Introduction:

I got a sense of what brightness and size work for notifications in Part1, and understood what concept and animation works best for participants in Part 2. This next research study was intended to put everything in the past research studies together and see if they all fit together. 

Part 3 Primary Research Questions:

• Can participants open notifications and interact with them comfortably? 

• Do participants feel like the flow of receiving and interacting with notifications is consistent with the rest of the UI?

Participants:

10 Mojo employees that were not a part of Part 1 or Part 2, but have interacted with this interface before.

5 people who are not Mojo employees and are interacting with this interface for the first time.

I wanted to test the interface with participants who had a variety of familiarity interacting with the interface. Getting a few participants from outside Mojo would give us insight on their first impressions and see how intuitive the interface is. 

Part 3 Methods:

• Usability test, semi-structured interviews

I made decisions to design the study in a way that would mimic a real world scenario users of the lens might face, while balancing the current tech constraints. 

Part 3 Results and Impact:

After this study, I compiled a document of proposed short-term decision changes and long-term concept changes. I prioritized these findings in a way that made sense for the future timeline of the lens interface design and milestones the design team was trying to reach. 

Overall Reflection:

I enjoyed conducting this multi-step research plan that built notifications up from a behavioral perception level to a fully designed and integrated feature. It paid off to “start from scratch” when creating this notification feature because it let the design team and I explore aspects of notifications we may not have otherwise questioned. This was a key decision I made with regards to the timeline of the Mojo Lens development, knowing that there was plenty of time before the notifications feature would be used on an actual lens interface; this allowed me the flexibility to explore this feature slowly and thoroughly. The research I conducted helped guide the design of this feature all the way from minor design recommendations to major concept shifts relating to how notifications relate to the rest of the UI. Through the process of designing and conducting these studies, I have learned how to advocate for what I believe to be the best next step for research and for major UI shifts that would better the user experience.

Looking back on the whole series of notification research I conducted, there are a few things I would have done differently. I would have conducted a slightly more rigid analysis in the behavioral study (Part1) and collected data in a way that could have been analyzed through an ANOVA. I was still able to gather data and synthesize findings from observing the way their eyes moved and other aspects of the eye-gaze data, though it was very time consuming and involved a lot of me creating assumptions. These assumptions were later validated in Part 2 and Part 3, however, I remember wanting to be more confident in the quantitative data I had collected. I would have been able to feel more secure in those findings if I had conducted a test that adhered to more of the standards of quantitative testing.