Loading page…
An AI fitness app that builds custom smart training plans for gym strength training
70K+ subscribers
He didn't dream up a category. He used the market-leading app (Fitbot) heavily, found its training plans “weird and dangerous,” then copied its UI/UX wholesale and swapped in a training engine that actually works.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Gravl's origin is refreshingly anti-formula. Julian didn't wake up one day wanting to build a fitness app — he first used Fitbot, the category-leading product, heavily, and found that the training plans it generated were “weird and dangerous”: unreasonable exercise pairings that could genuinely get someone hurt. That flaw, which he hit personally, was the opportunity itself. His reasoning was cool-headed. If the leading app has already proven that users will pay for AI fitness plans and that this UI/UX works, then the smart move isn't to invent from scratch — it's to copy the market-validated product form and be genuinely excellent at the one place it fails most dangerously: plan quality. In other words: clone the shell, swap the engine. The team started with Julian and two others and is now 13-14 people. They launched, added a subscription, and immediately started buying ads; today they do over $440K a month with 70K+ paying subscribers. What's rare is that Julian laid the whole cost structure out in the interview: ad spend runs about a third of revenue, Apple takes 15%, plus salaries for a dozen-plus people. That makes the $440K/mo unusually transparent — it isn't passive income, it's a hard business that welds product quality, paid acquisition, and creative output together.
The seed users came from a post that was not about fitness. Julian wrote up how he built the app, in real technical detail, and posted it in a developer community on Reddit; within hours it had several hundred upvotes and more than 300,000 impressions, and the first few thousand users arrived.
The quick-reference card — acquisition channels, replication playbook, and risk map — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself.
The app ran free first, long enough to confirm people would actually follow the plans, and then the paywall went in. Its placement is the mechanism.
Growth here does not spread on its own, it is bought, and what compounds is the ability to buy it cheaply. The loop is arithmetic: ad spend is held at about a third of revenue, so every month of subscriptions funds a larger month of buying, and the constraint is never budget but whether the creative still returns.
The first quiet stretch was the build: two to three months to an MVP, with the workout logger finished first and the hard part layered on after, meaning the engine that turns equipment, goals, frequency, sex, weight, age, and one-rep maxes into a week of safe programming.
The rest of this breakdown — 10 more sections — is in the Unicorn Club, with every section of all 1,335 ideas: first customers, first dollar, the flywheel, and the silent stretch.
Join the Unicorn Club — $5/moCancel anytime, in one click