Gravl · B2C · AI gym workout app
Copy a top app's design, but fix the part its users complain about
Founder says it gave the app its edge; no direct numbers
Workedpositioning-nichefeature-development
What they did
Dropped a me-too Hevy/Strong-style workout tracker that added no value. After a friend showed him Fitbod, he used it and researched it, finding its generated workouts weird and sometimes dangerous. Built 'Fitbod UI/UX with an actual proper workout engine' that accounts for equipment, weekly goal, gym frequency, consistency, gender, weight and age. MVP took 2-3 months.
What happened
Became the basis of an app now at $440K/month; no direct attribution.
- Stage
- Pre-launch (idea)
- Effort
- 2-3 month MVP
“But then, I started using it, and doing some research, I noticed that the workouts were weird, and even a bit dangerous sometimes, and like the more I used it, I was like, wow, okay, these workouts are actually bad. That's what triggered me.”
“We need to build Fitbod UI, UX with an actual proper workout engine.”
Related topics
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from this episode
- Gravl: Post a 'how I built it' thread on Reddit with the technical details300K impressions, a couple thousand users, and bug reports that proved demand
- Gravl: Start paid ads the day you add a subscription, once people want the appFirst subscriber within 10 minutes; ads are the main channel at ~1/3 of revenue
- Gravl: Translate the app to Spanish and run ads in South America for under $50 a dayWorked from the start and still does; US ads seen as too expensive
- Gravl: Pay creators ~$50 a video and copy what competitors run in Meta's ad library
- Gravl: Ask detailed training questions up front, then show a 'building your plan' screen
- Gravl: Put a hard paywall right after onboarding, before people even sign inFounder says 'a lot' pay before seeing the product; no rate given
- Gravl: Offer 24/7 in-app chat answered by a real person, not AIFounder says they grew a lot with it and users value it; no numbers