Four years building Zing Coach from an early-stage app into a product millions of people train with. Every design decision came down to one insight: users who completed four workouts in their first month stayed for the long term.
Period
Sep 2020 — Present
Role
Head of Design
Platforms
iOS / Android
Focus
Product Design, Retention UX, AI Integration
The challenge
Most fitness apps lose users in week one
People download with high motivation and quit before seeing results. The design problem wasn't onboarding or workouts in isolation — it was the gap between the promise of AI personalization and what the product actually delivered day-to-day.
Retention is not one decision
No single screen keeps people. It's the accumulation of small moments where a product either earns trust or loses it. Our job was to find those moments, test what moved the needle, and make the right call — even when it felt counterintuitive.
What we reached
40–55%
Month-1 retention
4.5M
Tracked workouts
2.2M
Downloads
4.8★
App Store rating
"This app is doing all the things for me. Being able to pick out your trainer and being able to adjust breaks and exercises is top tier. I also love the chat bot."
App Store review · ★★★★★ · "Very Pleased" · June 2026
"This keeps me more motivated than going to the gym. I hate working out in front of people, so I enjoy using this app from home — it made it easy to choose my workout type and the kind of exercises I want to do."
App Store review · ★★★★★ · June 2026
Retention vs. the category
Most fitness apps lose over 90% of users in the first month. The category average sits at 8–12% by day 30.
Zing CoachFitness app average
Month-1 retention is the confirmed figure; the surrounding curve illustrates the gap against the category.
The habit moment
Subscribers who completed four workouts in their first month kept training and kept their subscription. One number, validated across every user segment, became the target the whole first-month experience was designed around.
How we validated it
Not intuition — correlation analysis against long-term retention, defined as still working out nine weeks in. We tested workout counts, workout minutes, and content mixes; four completed workouts was the strongest, most consistent signal.
What it changed
Activation was redefined from "started a workout" to "reached the habit moment". The first month became a guided experience built to get every new subscriber to four workouts — including a dedicated 4-workouts-in-4-weeks habit feature.
Nine moments
Nine decisions that shaped how people experienced Zing Coach. Each came from data, testing, and millions of users.
Onboarding: More screens, better retention
"We tested everything shorter. Every single version underperformed. People needed the full explanation of how the AI worked."
Problem
Users didn't trust generic apps. They needed to understand how the AI personalized their plan or they'd quit.
Tested
Screen counts, question order, depth. Every shorter version had worse week-1 retention.
Shipped
A longer flow that explained the AI step-by-step. Conversational, not corporate.
Result
+27%
week-1 retention vs. shorter onboarding flows
Every time we tested a faster path, people left earlier. Investment in explanation drove commitment. We never built a "quick start" — every time we tried to shorten it, retention dropped. The instinct to simplify was wrong here.
Workout flow: Four workouts make the habit
"The data was clear: users who completed four workouts in their first month stayed for the long term. We rebuilt the early experience around that one insight."
Problem
A timer and rep counter felt lonely. Users completed sets but didn't feel coached. No reason to come back.
Tested
What made workouts feel coached: guided vs. unguided modes, real-time feedback, coaching cues, acknowledgment. Every unguided version dropped completion — and four workouts in month one was the retention threshold to hit.
Shipped
Real-time form feedback. Guided reps. Coaching cues during rest. Acknowledgment after each set.
Result
4 workouts
guided coaching moved more users through the first four — completion rose in every test vs. unguided modes
Subscribers who hit it stayed for the long term, validated across every user segment, so everything in the first month was designed to get people there. A simple, unguided mode stayed on the cut list: each experiment in that direction traded completion for convenience. Guidance costs resources — losing users costs more.
AI Coach: Unpredictable content needs flexible systems
"When the AI went live, it broke every fixed component. Text was 3× longer than we designed for. We had to rebuild everything."
Problem
Users needed coaching in the moment — mid-set, mid-rest. But AI answers were unpredictable: one sentence or a novel, impossible to act on during a workout.
Tested
A standalone chat surface vs. coaching embedded in the workout. One generic voice vs. personalized coaching tones.
Shipped
Chat embedded in workout context, answers shaped to be actionable mid-set — with the design system rebuilt on flexible min/max rules so any answer stays readable.
Result
+13.6%
messages per user with a personalized coaching tone
Users engaged more when the AI spoke to them as individuals, not when it buried guidance under caveats. So we rewrote the coaching voice from hedged to clear: "This might work for you" read as uncertainty, while a real coach commits to a recommendation, then adapts as they learn more about you. Directive language moved people to act — and the AI kept adjusting from their feedback. It stopped narrating its own doubt, not its honesty.
"Body composition changes take weeks. People need to see progress now or they quit. We made invisible change visible."
Problem
Real fitness progress (muscle, body comp) takes weeks. Users quit because they don't see results in day 3.
Tested
Weekly body scans with video comparison, personalized progress plans, activity streaks, and strength progression — measuring which drove the fastest perceived progress.
Shipped
Body composition tracking with visual before/after, a personalized workout and progress plan refreshed weekly, activity streaks, and a Strength Score metric. Made the invisible visible without lying.
Result
0.5M
body scans completed — progress made visible at scale
60% vs 18%
viewed their Strength Score after two workouts — test vs. control at rollout
Making invisible progress visible moved people from frustration to motivation — they stayed because they could see themselves improving. Body scans require hardware, and we didn't build a lower-friction option; the friction of accurate progress tracking was worth it for the engagement it kept.
Fitness test: Reframe data collection as personalization
"The same exact questions were friction when framed as assessment, but personalization when framed as 'learn about yourself.'"
Problem
Users treated the fitness test like an exam — something to get in shape for first, then pass. So they skipped it, and we lost the data that personalized their plans.
Tested
Copy, timing, and positioning — "assessment" vs. "personalization" framing, and at which moment in the journey to ask.
Shipped
Reframed as "Let's learn what works for you" and surfaced at the moment it clearly powered their plan — not as a gate to pass. Same data, different story.
Result
78%
fitness assessment completion after reframing it as personalization, not evaluation
The same questions became welcoming instead of intimidating. Shortening was never the fix — the full assessment fed the personalization, so the friction had to be solved by framing and timing, not by cutting questions.
Injury prevention: When to coach instead of correct
"Telling someone their form is wrong mid-workout kills momentum. Coaching them through it keeps them moving and safe."
Problem
Early-stage users have bad form. Correcting them creates friction. But bad form causes injury and dropout.
Tested
Real-time correction vs. coaching through good form — measured on engagement and injury-related drop-off.
Shipped
Coaching cues that guide form without stopping the workout. Flexible movements, proper breathing, safe progressions.
Result
Day 14 → 30
injury-related drop-off pushed later by coaching through safe movement
Coaching through safe movement worked better than stopping people mid-exercise to correct form. Coaching is harder to build than correction, but we invested in real coaching language, not just error messages.
Nutrition integration: Coaching engagement separate from workout engagement
"People who track nutrition with AI coaching stayed longer than people who just worked out. But these are separate habits."
Problem
Nutrition matters for fitness results, but adding it early overwhelms users — the challenge was finding the right moment to introduce it.
Tested
Nutrition in onboarding vs. after first 10 workouts. Timing, copy, and AI coach integration.
Shipped
Nutrition after they've built the workout habit. AI coach who understands their specific goals and preferences.
Result
2×
better 90-day retention among nutrition-tracking users
Introducing nutrition after workout-habit formation meant people actually stuck with both. We delayed it to focus on the workout habit first, which meant a smaller initial feature set and slower time-to-value for people who care about diet.
Notifications: Consistency beats novelty
"We tested fancy AI-personalized notifications. The simple ‘it's time to work out' message had 2× the engagement."
Problem
Notifications become noise — most apps train users to ignore them. Ours had to stay useful without becoming annoying.
Tested
Frequency, timing, AI personalization, copy. Simple vs. clever. Scheduled vs. adaptive.
Shipped
Consistent daily notification at their preferred time, proactive morning check-ins from the coach, plus a personalized weekly summary — what they achieved, how it moved their progress, and the plan for the week ahead.
Result
+28%
opens vs. clever, varied messages
+19%
session conversion
+8.1%
app opens within 6h of a morning coach check-in — stat. significant
+7.9%
workouts completed within 24h of a check-in — stat. significant
Predictability beat novelty. We didn't personalize notification content deeply — consistency mattered more, and people knew when to expect a message and trusted it was relevant.
Daily Tasks: Retention that compounds
"Week one showed almost nothing. By week four the uplift was undeniable — the effect compounded as the habit loops took hold."
Problem
The AI coach lifted motivation, but engagement across the rest of the product stayed uneven — users lacked a daily, personal reason to open the app beyond the workout itself.
Tested
A/B on ~19,400 users: a personalized Daily Tasks feed on the Home tab — coach check-in, muscle readiness, tests, and progress cards — against the existing static section.
Shipped
The Daily Tasks feed, refreshed once a day, rolled out to all users — the foundation for personalization at scale.
The effect was invisible in week one and undeniable by week four — retention compounded as the habit loops took hold. Decomposition showed exactly where it came from: stronger progress feedback loops through Body Scan and more conversations with the coach. Tests we didn't touch stayed flat, confirming the mechanism rather than a general novelty effect.
What this taught me
Retention isn't one big feature. It's a thousand small choices about whether to add friction for better results, or remove it for better adoption. Sometimes the answer changes week-to-week as people move through the habit-formation journey.
The hardest part isn't design or engineering. It's staying disciplined about what matters: does this choice actually improve retention, or is it just polish? I said no to a hundred features to say yes to the right nine decisions.
