The original is one click away. Open original ↗
How a Developer Built a $400K/Month AI Fitness App in Two Months
Executive overview
Julian Gargicevich spotted a popular fitness app with bad workouts and built a better version in two months. He launched on Reddit, found immediate traction, then scaled with paid ads and Spanish-language markets.
The result: 70,000+ subscribers and $440K in a single month.
The core insight: copying a competitor's UI while fixing its core product failure is a faster path to product-market fit than building from scratch.
Finding the idea
- Discovered Fitbot — a popular workout app — through a friend
- Noticed the generated workouts were poor and sometimes dangerous
- Decided to replicate Fitbot's UX but build a proper workout engine underneath
- Previous startup had drained energy without product ownership; this was the reset
Building the MVP
- Initial MVP took two to three months
- First phase was a basic workout tracker; second phase added AI-generated custom workouts
- Key complexity: combining equipment, goals, frequency, gender, weight, and age into coherent plans
- Built with React Native / Expo (frontend) and .NET (backend)
- Uses Cursor for AI-assisted coding, but keeps tight control over which files the AI touches
Launching and finding product-market fit
- Posted a technical "how I built it" thread on Reddit under the original name Gains AI
- Got hundreds of upvotes and 300,000 impressions within hours
- First thousand-plus users came directly from that single post
- Users were developers who also gym — gave immediate bug reports and feature requests
- App was free at launch; early feedback was the signal to turn it into a real business
Scaling to $400K/month
- Added a subscription model and ran first paid ad — got a subscriber within 10 minutes
- Translated the app into Spanish and ran ads in South America at under $50/day
- Lower competition and ad costs in Spanish-language markets drove early growth cheaply
- Team grew from 3 founders to 13–14 people (mix of full-time, part-time, contractors)
Cost structure
- Paid ads (Meta, TikTok, some Google and Apple Search): roughly one-third of revenue
- Salaries: $50K–$80K/month
- Apple's 15% revenue cut is a fixed overhead
- Infrastructure (servers, AI tooling): ~$1K/month
- Mobile measurement partners (MMPs): ~$1K/month
App design decisions
- Hard paywall shown before users see their generated plan — converts strongly
- Onboarding asks detailed questions (split, equipment, goals, experience, excluded muscles) to personalise from the start
- UGC (user-generated content) is the primary ad format; cheap to produce and easy to test
- 24/7 in-app support chat staffed by real people — cited as a key retention driver
Tips for building consumer apps in 2025
- Validate willingness to pay before spending on ads — not just that the product works
- Target cheaper ad markets first; Spanish-language markets cost far less than the US
- Start small with influencers — $50 for a piece of content is enough to test
- UGC ads outperform polished creative; AI tools and CapCut make them fast to produce
- Use Meta Ads Library to study what competitors are running and copy what's working
More like this — when you're ready for early access.
Join the waitlist for a personal account and content recommendations based on what you're working on.
No spam. Unsubscribe at any time.
You're on the list. We'll be in touch before launch.