Range of Motion Tracking
Motivating users to squat deeper to get more out of each rep.
Range of motion tracking is an important part of many strength training movements, ensuring you’re using all your muscles efficiently.
Form feedback vs. Form guidance
With Tempo’s mission to be a personal trainer, our members look for real-time guidance to know if they’re doing the right things in a movement, working hard enough, and breaking form in any harmful way. In Tempo’s first 2 years of a product, “form guidance” was only visible in limited “form feedback” forms, only triggering when members did something wrong, and when they corrected it. Because of feedbacks were only triggered around 10% of the time, and it minimized the value of Tempo’s “coaching” because it was invisible every other time.
How might we add value to the in-class experience, providing more focus and guidance to members, not just when things are going “wrong”, but also to reassure them of correct behaviors and form?
Messaging experience for form feedback on the Tempo Move.
Process and my impact
The “magic” of Tempo’s 3D vision, was something users experience as each rep was counted, and when a user did something wrong, triggering “form feedback.” But there were limited exercises those feedbacks were tied too, so we knew we wanted to create something that felt more real-time, and could provide focus all the time.
For this project, I worked with one other visual designer, and spent a lot of time consulting our director of fitness and our CTO, with product support from 2 product managers overseeing both the computer vision needs and the user experience needs.
Once we aligned on a real-time range of motion guidance feature, my role was to determine how it would work, when it would show up, and design, spec and test all the variants and states we’d need—all within 6 weeks.
Considerations
Just to show a bit of everything we considered, there were three categories of key questions we tried to answer and explore—how we build this and collaborate with engineering, how this needs to fit into the UX .
From the process side:
How can we design and iterate quickly in order to get stakeholder alignment, but also unblock any computer vision work as quickly as possible?
I worked with a visual designer to do a wide exploration of the meter, both in a horizontal or vertical format.
From the fitness side:
What are the different states we were going to need?
What information is most important to show to users to change their positively behavior in a class?
What level of variation might be need for different types of members or different types of strength training exercises?
I focused on diagraming out the different states and different exercises.
From the design side:
How can we provide more focus to the workout experience so that the guidance is actually changing how the user works out, vs adding to noise and clutter on the screen?
When does this range of motion feedback matter and when is it valuable?
In the above gallery, I wanted to consider how the Range of Motion meter could fit into the various workout layouts—Move and (shown) Studio. I thought through how we could coach to it, and how the real-time ness and any messaging / feedback would help users understand how to use it.
Building and iterating on the MVP
To have the most amount of impact and test success first, we started with squat movements across Move and Studio users. As we designed and reviewed with stakeholders, we built prototype versions and watched user’s work out, try it out and then iterated on small details, better understood which exercises worked well and didn’t, and ironed out kinks and bugs.
Spec and handoff
We needed states for active and rest states, and then wanted to make sure when we handed off to engineering, all the zones, behaviors, and states were expected and clear.
Learnings and community reaction
From the experiment on studio, we did see that members who saw the ROM meter hit an average max depth of 95% over the control of 85%.
While it was cool to see our UI change what members focused on in the workout, we learned we really needed to make the range values customizable or allow them to turn it off if they have any injuries, etc. The next phase of this work was to add it to more exercises, and show users change or improvements over time.