Windy City Lab is developing an internet-connected smart mirror for consumers called Glance. We were tasked with researching the marketability, usability, and feasibility of the range of components of the product. Our focus was split between the interface and functionality of the mirror itself, and the information architecture and wireframing of a companion app used for the mirror’s configuration. Through research, usability testing, prototyping, and wireframing, we presented a viable and unique product to our client in just over two weeks.
Our initial prompt was to provide added value to the bathroom experience through a mirror. However, we decided to investigate use cases for Glance beyond the morning bathroom setting as well. We had free range to explore the industry as a whole, so through research we aimed to discover unique opportunities for the concept.
Our client intended to implement unique user ID through facial recognition software. This became a huge privacy concern, as we later proved.
One of our primary goals was to discover if Glance was desired by users, what problems it helped them solve, and what goals it helped them accomplish. We also wanted to explore the extended marketplace of smart mirrors, and their various applications in daily life.
We looked at 17 competitors, both direct and indirect, to see how they coped with similar challenges that Glance would eventually face.
We quickly noticed the clear opportunity to differentiate Glance through a unique combination of glanceable information with unique user identification.
We interviewed high functioning executives, parents, techies, and luxury consumers for user behavioral insight, and how they might interact with the product. We found that privacy was, not surprisingly, their primary concern.
Put simply, our core findings found that people need information in the morning to plan for their day. This is typically presented on multiple devices which disengage users from their environment, and is nested in multiple apps, which decreases efficiency.
To define our target user, we learned about their motivators, frustrations, and usage habits with their existing mirrors. With the prevalence of different types of mirrors and the use cases among different types of people, the challenge was to hone in on a tool that could be used universally.
We had our interviewees walk us through their morning and evening routines, and where they source their information throughout the day. Our resulting personas identified three distinct users.
We tracked cognitive capacity because of the relevancy of mental effort required to learn new information as users prepare for their day. We tracked three separate days for our primary persona, Angelina.
Through this exercise, we learned that people visit bathrooms multiple times a morning with the ability to handle different cognitive loads at different times. We noticed clusters in the morning and evenings with gaps in the afternoon.
People were checking their phones, laptops, and tablets in the morning to check for messages, communicate, check weather, monitor commute time, and mentally prepare for the day. This deeper understanding of mirror usage illuminated Glance’s value propositions in users’ daily routines.
Knowing this, we established a set of design principles to guide our thinking going forwards. Our synthesis culminated in six core values:
Glanceable: Information will be easily digestable at a quick glance through concise and communicative language
Delightful: Inspired feelings of joy through friendly feedback. A simple and clean interface with subtle interactions
Personalized: Information and how it's displayed can be customized by and for the user
Unobtrusive: Becomes part of the user's natural home environment and won't disrupt them from their daily routine
Interconnected: Facilitates information that helps the user meet their goals
Trusted: A secure platform with accurate, reliable, and timely information
Our goal in ideation was to brainstorm the entire range of potential features, and funnel down to a minimal viable product. The biggest hurdle was to address the privacy concern posed by the camera.
We initially thought about moving the mirror out of the bathroom space, to allay privacy concerns. However, we found that users were either still concerned about a camera in their homes or that mirrors weren’t looked at as frequently elsewhere. We thought about different forms of interactivity to substitute facial recognition. Voice, motion, gesture, and touch were all tested by users later on. We interviewed an AR/VR SME to explore the feasibility of gesture technology.
For customization, users would be able to set different mirror views for different parts of the day, and rearrange the widgets to their liking. We recognized the risk of overwhelming the user with new and complex technologies, and sought to achieve a minimalist solution going forward.
The most important task flow to think through was the setup and onboarding flow for a first time user. Because of the steps needed to sync the mirror to the phone, input user biometrics, and teach a completely new technology, we strived for simplicity. We looked at comparable products that required similar user inputs, and drew inspiration from Nest’s Dropcam. We created a simple linear flow, with one action per screen. To train the user, we forced them to interact with Glance before completing the setup.
We also prototyped wireframes for the mirror to get an idea of proportions and spacing. We paper prototyped our first round of testing, and later graduated to a mid-fidelity wireframe.
Our goals for testing were to determine how users engage in customization, interactivity, and integration with outside apps, and user identification. We tested men and women, and compared facial recognition against voice ID. We also tested the product in several settings to gauge legibility in different lighting environments.
Primary information was placed at the top of the mirror, while secondary and more detailed information was placed towards the bottom.
Customization: Users wanted Glance to have a personality, and less information was desired in the evening
Interactivity: 100% of users expected to be able to interact with Glance, and users wanted to drill down to desired content
Integration: All users wanted a music feature, and the portability of bookmarking and saving content to their phones
User ID: Users were uncomfortable with facial recognition technology, and were naturally inclined to speak to Glance
Given our testing results, we presented the benefits of discarding the facial recognition technology for a more trustworthy voice ID system to our client. This versatile solution addressed several of our key insights simultaneously. Coupling voice control with an expandable form of content lent itself to the clean aesthetic of Glance, while appealing to users’ desire to drill down for detail.
Users also completed a mirror-facing onboarding where they were introduced to five distinct voice command functions. Users learned basic commands that they could use to control widgets while Glance simultaneously learned the users’ voice to uniquely ID that person for future reference.
Challenges to consider with this solution would be to create a thorough yet friendly onboarding to educate users about commands, and ensuring that users accept a new form of interaction that is not as fast as touch.
Advantages of this solution are considerable:
- Dispels nearly all privacy concerns
- No fingerprints on mirror
- Hands free is great for a wet environment
- User-friendly for the disabled
- Onboarding is FUN
- Opportunity for customization and personality
- People are more comfortable with speaking to tech in privacy of their own bathroom
Our final recommended MVP featured the following:
Our testing validated our core problem statement in that people confirmed it would make their morning routines much more efficient. Furthermore, people would be more engaged with their surrounding environment by reducing morning screen times, as Glance would serve as a replacement for their phones. Users were satisfied that they could get all of their information in one consolidated place. Last but not least, users had no privacy concerns whatsoever with voice functionality. We ultimately struck a balance between glanceability and satisfactory content, while giving users a sense of control. Our research, findings, and client deliverables are currently being used to pitch investors.