Prevayl

FitnessCheck™

A guided workout to give a picture of your current health and fitness, and an activity you can use to measure cardio fitness progress.

Goals

Using a high accuracy fitness tracker paired with a phone to develop a maximal fitness test, which can be picked up and followed without outside instruction.

Scope

Create a new app feature for a treadmill fitness test, displaying instructions and current heart rate metrics during the test, and communicate the user's fitness levels in a summary after the workout.

The result of this test can then be used to determine the user's heart rate training zones and maximum heart rate, instead of estimates, enabling more accurate fitness insights.

Requirements

First was understanding how the test works, and talking with the data scientists who devised the test to understand how it would function under the hood.

The primary metrics that were being calculated through the test were VO₂ max (maximal oxygen uptake estimate), heart rate recovery (how much your heart rate reduces after intense exercise), and the peak heart rate achieved during the test.

As these results are partly hereditary, based on age, or based on sex, its important that results are compared against official reference ranges for people of the same age and sex as the user, or compare results against the user's previous results to show personal progress.

Research

This type of test is typically instructed by a trainer or health professional, so it was important that the app communicated to the user in a similar way to enable them to do the test on their own.

• I observed a number of tests being conducted by a trainer in the on-site gym, to get an idea of how trainers communicate the instructions to the participant.
• I recorded the ways the trainer would communicate the test, the types of reinforcements, and the cues they would give to the participant at each stage.
• I recorded the participant's reactions to the test, and whether they changed their pace in line with the instructions.

It's crucial that these instructions are followed to get accurate results from the test, so a draft set of instructions were created, mimicing a trainer, that would be used in the initial round of testing the UI. It's crucial that these instructions are followed closely to get accurate results from the test, so I added the initial instructions to a prototype for testing.

Prototyping

I created a simple layout that could be reused for each phase of the test, with instructions that update depending on the phase.

The goal was to do a user test with a Figma prototype to see if the instructions could be followed without the trainer.

Test observations

Observation: Heart rate was being focused on as the central part of the screen over the instructions and time left. Determining the user's maximum heart rate was an important part of the feature, but the time remaining and instructions needed more importance.

Iteration: Redesign the layout to give more prominence to time and instruction.

Observation: We needed the user to ramp up their intensity in a specific way, and a single instruction per phase was not enough during the test.

Iteration: Add multiple prompts per phase that update as time progresses. Rather than a single 'Warmup' instruction, the Warmup phase was given 6 prompts that build up their pace gradually. Each phase was given 1 prompt per 30seconds.

Observation: The phone was typically placed on the treadmill, and some users struggled to see details when running.

Iteration: The plan was to have audio instructions for the test, however this was out of scope for the first release, so the UI would be made bigger and bolder for release and audio would be worked on for future iterations.

I made adjustments based on this feedback and retested. The results with this version were much better, with all test participants able to complete the test without the trainer.

This was still a low-fidelity Figma prototype and lacked some of the functionality of the final feature, so I started work with developers to design and implement the final product.

I put together the high-fidelity designs, along with a color scheme for the different types of instructions and feedback.

More testing

An early version of the feature was built so it was time for testing with the real thing.

I created a test plan with the following goals:

• See how easy/difficult it was for the participant to find and begin the feature.
• Observe how participants interact with the feature, before, during and after.
• Do they understand the purpose of the feature?
• Is the test completed accurately?
• Does the participant reach close-to or above their maximum heart rate, as the test requires?
• Do they understand their results?

Results

The results of the test showed participants found the feature reliably.

They all read the feature introduction screen without instruction, selected the correct gym equipment, and started the test.

A few participants did not ramp up their workout intensity as expected and commented that they could have pushed harder in the high intensity part of the workout. Domain experts noted to rewrite some instructions for future tests.

Some participants wanted the instructions to respond dynamically to their current heart rate, so this was noted for future iterations.

All participants read and understood what the results were communicating to them.

After a round of text changes, tests were performed again with new people. All participants reached a peak heart rate within the expected range for the test, and a number of participants set a new record maximum heart rate during the test.

The feature was then released to the public as part of the Prevayl app.