The Story
Julian is a developer from Argentina who built Gravl - an AI fitness app that provides smart, personalized gym workouts. The app now has 70,000+ subscribers and made over $440,000 last month.
Julian grew up around fitness - his dad owned a fitness center. He got into software engineering, moved to Australia, and worked at companies from small startups to Atlassian. During COVID, he and his partners started an influencer marketing platform for mobile games, which evolved into a marketing agency. Not what they signed up for, but they learned a lot about user acquisition and saw the revenue numbers fitness apps were making.
The turning point came when a friend showed him Fitbot, one of the biggest workout apps. Julian thought it was amazing at first - it provided workouts on the spot without much effort. But as he used it more, he noticed the workouts were weird, sometimes even dangerous. That's when it clicked: "We need to build Fitbot's UI/UX with an actual proper workout engine."
The initial MVP took 2-3 months. For validation, Julian posted on Reddit about how he built the app (then called Games AI). The post got hundreds of likes, 300,000+ impressions, and thousands of users. Many were developers who found the gym intimidating - perfect early adopters who provided bug reports and feature requests.
Once they validated product-market fit, they added subscriptions and started running Meta ads. They got their first subscription within 10 minutes of activating ads. They also translated to Spanish and ran cheaper ads in South America to test and learn.
Key Insights
The "Clone Better" Strategy
- •Find an app that's crushing it but has obvious flaws
- •Don't just copy - identify what's actually broken
- •Build the same great UX with a better core product
- •Julian's insight: Fitbot had great UI but bad/dangerous workout logic
Reddit Launch Strategy
- •Post in relevant subreddits about HOW you built it (technical angle)
- •Developers make great early adopters - they give detailed feedback
- •Use Reddit to validate before spending money on ads
- •The post got 300K+ impressions and thousands of first users
Paid Ads Playbook
- •Start small ($50/day or less)
- •Test in cheaper markets first (South America vs. US)
- •UGC content is king - most successful ads are user-generated
- •Use AI tools to create test videos cheaply
- •Copy competitors - Meta Ads Library shows what's working for them
5 Tips for Building Mobile Apps
- •Validate before spending on ads - not just that it works, but that people will PAY
- •Start in cheaper markets to learn before scaling to US
- •Start small with content - $50 influencer posts, AI-generated videos, CapCut edits
- •UGC is king - real people using your app converts best
- •Copy what works - Meta Ads Library is public, study your competitors
Tech Stack
- •React Native + Expo (mobile)
- •.NET (backend core functionality)
- •Next + React (internal admin dashboards)
- •Cursor (AI coding - very specific about which files to touch)
- •24/7 human support chat (not AI - users value real responses)
Costs & Margins
- •Meta/TikTok/Google/Apple ads: ~1/3 of revenue
- •Team salaries (13-14 people): $50-80K/month
- •Apple's 15% revenue cut
- •MMP tools: ~$1K/month
- •Infrastructure (servers, AI, tools): ~$1K/month
Advice
- •Be proud of what you build - motivation from loving your work can't be matched
- •Keep pushing, don't give up too soon - building is not even half of it
- •Be ready for punches - ads that don't work, bugs, setbacks
- •But also know when to give up - don't drag a failing product for an extra year
Resources
- •Follow Julian: https://x.com/julian_gargi