Project Details
Tools: Sketch, After Effects
Role: Design
After 7 years the Fitbit App dashboard was in need of a refresh. Our goal was to break the dashboard down into customizable tiles. Each feature could get summarized into that tile for glanceability. Tapping on a tile would let you dive deeper into the data.
This project was a very big step for Fitbit. We relied heavily on prototyping and A/B testing with a variety of users. We used this data to help refine our updated dashboard over the course of the project. Making specific design decisions where we saw added value from our UX tests.
Dashboard Tiles
Why change the dashboard?
- Dashboard hasn’t changed much in 5 years
- Design doesn’t effectively support our different users
- Visual design is well-liked but is utilitarian
- Doesn’t surface or provide quick access to popular features
- Needs to adapt to varying screen sizes
- Needs to accommodate new features
- Can the dashboard drive interactive engagement?
Dashboard through the years.
Quick Actions
Research
- When + why are people using the dashboard over the course of a day?
- What information do they care most about (and why)?
- What’s working and not working with the current design?
Methods
- Mobile Ethnography: 60 users journaled usage over a week
- Participatory Design: 6 in-depth interviews to explore mental models
- Card Sort: 150 respondents sorted items by category and relative priority
How Research affected our goals
- Consolidate activity & food tiles
- Make the dashboard more time-relevant
- Provide better access to key features (silent alarms, MobileRun)
- Make dashboard more visual
- Keep it simple
Dashboard - Motion Explorations
- Dashboard Pull to Refresh
- Dashboard Tiles
- Goal Celebrations
A/B Testing
- Dashboard is critical part of the mobile experience - we wanted to be confident before making such a significant change.
- Dashboard changes are always hard because users hate change.
- Understand how changes affect usage quantitatively.
- Release elements of the new design with new users.
- Release multiple variants such as tile order.
- Validate hypotheses.
- Impact to other parts of the app.
- Surprising results.
How we tested
- A/B tests in production iOS app
- 150,000 new users over 30 days
- Charge HR, Flex, Charge, Surge, One
- Usage of features
- Engagement and Retention
- Satisfaction
Results
Coming back more
- App sessions per week increased between 1-5% with the best variant
- 30 day daily retention up 1.5% → could amount to 1.5% more IDAU active on day 30 (~170k users)
Using more features
- More people discovering features: ~20% increase for MobileRun, Food and Water logging, 8% increase for Silent Alarms
- More frequent visits to fitness feature sections: ~5% increase to HR, Exercise, Sleep, Weight
- More people sending friend requests by 3%, more users in Challenges by 4%
Further Exploration
- Fewer people tapped into Mighty Tile stats and HR tile
- Tile placement has big impact on daily feature engagement KPIs
feedback
- CS contacts and Community posts are mostly positive
- Tweets and Facebook posts from users who want the new dashboard
“My dashboard went from having the large square tiles to having the small rectangle tiles tonight and I HATE it!!!! The format was much easier and I could see all five of my goals easily at the top. Now I have to search through the tiles to find my goals and I no longer get my screen to light up green when I hit all my goals. This new format sucks!!! Bring back my dashboard, or at least allow us to have an option as to how we want our dashboard set up.”
“I had the new version (big tiles with circles) and I loved it! Then suddenly today it changed? Will we be able to change it once this test is done? Because I can't stand the bar/list style.”
“Please please please bring back the test version. Old one is dreadful. Test version was much better design and less scrolling. Truly a ‘dashboard’ layout. This old crap should be called ‘endless scrolling list’ not dashboard.”