I audited 6 fitness apps, studied sobriety apps (which hold onto users far longer than fitness apps do), and ran 6 in-person interviews across three cohorts. The participants hit the same friction zones whether they were lapsed, active, or new to fitness apps.
Project phases
Research
Apr 8
Empathy & Definition
Apr 22
Ideation
May 3
Design
May 12
Implementation
May 19
01
01 · The Problem
Most people quit. The app lets them.
Research context · Retention data
71%
of users abandon their fitness app within 3 months
90%
quit their fitness goals entirely after three months
Dr. Muscle (2024) · StriveCloud (2024)
"Many people begin a workout regimen but lose focus and motivation after a short amount of time. Conventional gym apps frequently lack elements for goal setting and habit building."HabitaBull Lifting · Project Brief, Fall 2023
Fig. 01: Problem framing
02
02 · Research + Insights
The same friction, everywhere.
App critique. I read 6 apps for how they hold users, where they overcomplicate, and what they leave out.
Cross-industry research. I Am Sober: CBT-based motivation loops applied to fitness habit design.
Habit building framework. 6 principles from Atomic Habits translated to app features: identify, schedule, flex, set a goal, repeat, reward.
Interview participants. 6 people, Vancouver. Formerly used / currently active / never used: different question branches per cohort.
Fig. 02: Insight clustering across all 6 participants
What they wanted
- Simple interfaces usable mid-workout
- Progress visible over time, not just today's session
- Community or social layer for accountability
- Workouts that adapt to level and available equipment
What pushed them out
- Logging too time-consuming to maintain mid-set
- Apps felt generic, same routines regardless of goals
- No guidance for beginners on form, pacing, or rest
- Too many ads, paywalls, features unrelated to lifting
Fig. 03: Top 6 findings from interview clustering
Core design thesis: habit infrastructure is the product
03
03 · Design Process
From how might we to lo-fi.
I wrote six "How Might We" questions, one per top finding, to set the design direction. Then I mapped user flows for the two core journeys (building a custom routine and using auto-generate) and tested lo-fi prototypes against four tasks in Figma.
HMW 01Make it easy for new users to navigate through the app?
HMW 02Ensure that new users feel confident and informed when starting their fitness journey?
HMW 03Incorporate community support within the interface?
HMW 04Motivate youth to be disciplined and create healthy habits?
HMW 05Incorporate additional features and ensure a more engaging and supportive platform?
HMW 06Cater specifically to individual fitness goals and needs?
Fig. 04: Opportunities per finding, and feature ideation (green notes = built in Figma)
User Flow A. Creating a new lower-body workout routine: muscle group selection, exercise picks, animation previews.
User Flow B. Auto-generate: inputs equipment, target muscles, and available time. Outputs an editable routine with sharing and Spotify integration.
Lo-fi: Account setup. Name, past weights, favourite exercises, preferred workout days. Sets the "why" before day one.
Lo-fi: Routine builder. Full flow from profile to goal display to muscle group to exercise selection, minimal taps.
Lo-fi: Auto-generate. Equipment → body part → duration → editable generated routine. Lower barrier for beginners.
Lo-fi: Streak calendar. Workout calendar with a 6-day streak indicator and a photo album, so progress shows up as pictures, not only numbers.
Lo-fi: Leaderboard. Community competition by exercise (Bench Press, Shoulder, Back Squat). Congrats when you beat a PR.
Lo-fi: Exercise tutorial. Animated demonstration plus a list of common mistakes, so beginners know what to watch for.
04
04 · Brand + Solution
A bull with a grin. A system that sticks.
I built the brand around the mascot: a grinning grey bull who carries the HabitaBull voice. He's strong and a little stubborn, and he's on your side. Friendly without going soft, energetic without shouting at you.
6
Core features, one per research finding, each aimed at a specific drop-off point
3
Habit-building frameworks integrated: CBT loops from sobriety apps, Atomic Habits principles, autonomy theory
2
Core user flows prototyped and lo-fi tested with the target audience before I started visual design
0
Generic workout templates: all routines are either user-built or personalized via auto-generate inputs
Fig. 05: Brand guide and high-fidelity screens: goal tracking + community leaderboard
05
05 · Reflection
What I'd do differently.
Cross-industry research changed everything.
Studying sobriety apps was an odd move for a fitness project. But I Am Sober had already cracked the retention problem I was working on; it just talked about it in recovery terms. Looking outside fitness showed me fixes the six-app audit never would have.
The mascot did real UX work.
The bull started as decoration. By the end he was doing the work: he gives the app a voice, a reason to celebrate a PR, and enough presence that skipping a workout feels like letting someone down.
Lo-fi testing changed the hierarchy.
My first prototype opened with goal-setting, which felt logical and tested badly. People wanted to log a workout first and set goals later, once they'd seen what the app could track. I had built from the product's logic instead of from the first thing you actually want to do.