Google Nest Hub 2 with sleep sensing

Duration: Oct 2019 - Oct 2020 | Role: Lead product designer

Google Nest Hub 2 helps you understand your sleep and develop good sleep habbits by providing you sleep summary and personalized suggestions near your bedside. I Led design of sleep tracking and sleep coaching on the nest hub 2 while collaborating with teams from Google Fit, Google Home and Google Assistant. That means handling with: Design sprints, roadshows with 4 engineering teams, many executive reviews, rounds of user testing and iterations, coordinations with agencies...

01. Background

Sleep impacts every facet of daily life and is critical for long-term health. Quality of sleep is becoming the #1 concern for adults when it comes to health and wellness at home.

Teams at Google Nest Hub have been exploring ways to make their smart display more helpful. They know people already come to Google for information and tools to help them live healthier, happier lives, and they’ve specifically noticed more and more questions about sleep, exercise and health.

While my team at Google Health has been exploring ways to detect peopole’s sleep pattern using the latest Soli technology. In the early 2019, these two teams found each other and decided to bring the concept of sleep tracking and coaching to the second Nest Hub as they noticed a lot people already put their Nest Hub in the bedroom. When we look at the market today, there are lots limitations for sleep tracking and coaching. It's not accurate, not comfortable to track and quite complex to understand.

02. How to measure sleep? Not duration!

I was brought to the team in late Oct 2020, and was asked to create a sleep experience on a bedside device whose detailed data will be stored in Google Fit. Nobody knows what sleep experience should look like on a smart display as there is no primitive existence in the world.
To kick off the project, I first looked at all the sleep designs and research Google Fit has done for their app. I also mapped out all the potential moments users are likely to interact with this bedside device. According to their card-sorting user research, we know that users are very interested in the following information:

Based on my conversations with Sleep scientist Logan Schineider and readings on the book “Why we sleep”, I got to know that duration is not enough to measure a good night sleep. People will feel more rested when their sleep cycle matches their circadian rhythm. Also, Nest Hub has some other unique capabilities like accurately capturing when you went to bed and when you fell asleep and external environmental factors that disturbed your night.

Combining these three knowledge sources, I created the first version of sleep on Nest Hub and guided UXRers to validate the concept of masterclock. This concept of “masterclock alignment" resonates so well with users that it became the p0 feature of the product.

03. Redefine principles for a bedside device

The visual representation of sleep data on Nest Hub is ok but not great. We also dicovered in our research that some terms we provides in the product are too clinical and we should not simply align on onboarding tips to help users understand. From Jan - Mar, I worked with data visualization expert Alan McLean to come up with a more iconic representation of sleep. The interesting thing is to recognize the difference between a beside device and Google Fit app.

Based on these principles, we worked with Logan and Assistant visual team to break down the concept of masterclock into more understandable and actionable metrics. These metrics are tested very well with users. We broke sleep into Duration: How long you slept, Schedule: How consistently you go to bed and get out of bed, Quality: How restful your sleep was

04. Iterate based on principles and UXR

The refined principles really guided all the iterations and decisions, it impacts how I design data charts on Nest Hub to illustration styles we chose.

For example, for quality, I used to use actual number to identify how many wakeups and movements users had during their sleep, which triggered users to count near their bedside. So the final design hides the quantitative part of disturbances to avoid these behaviours.

For users to set up their sleep sensing, they need to lie in bed and calibrate their device first. The original design is quite sharp and feels like a medical device, based on the principles and aligned circle concept, we modified the visual representation to be more soft.

For sleep coaching, instead of showing bar chart on a bedside device, we focus more on proactive suggestoins and raise users awareness of their problems.

05. Drive alignment with LON, MTV, TYO teams

As all detailed sleep data from Nest hub needs to be stored in Google Fit while Google Assistant will be the transimitter for these two devices, I have to work with teams from London and Tokyo. Things I Learned from working with teams in different time zones is that Google Meet is always better than emails and chat. Personal collaboration is always better better than Google Meet. I organized a 1-week design sprint with Fit Design tam in London to align with sleep metrics and sleep coaching which accelerated the program critically. One of the key princple we aligned is the weight we put for sleep coaching on each platform.