Founder Playbook · Starter Story
12 tactics from Julian
How I Built It: $400K/Month Mobile App (Gravl)
Watch the full episode“I started using it, doing some research, I noticed that the workouts were weird and even a bit dangerous sometimes… Okay, this is it: we need to build [the incumbent's] UI/UX with an actual proper workout engine.”
Clone the category leader's UX — then fix the broken core nobody else noticed
Pure copycat builds add no value, but cloning a beloved interface while fixing what's actually broken underneath is a viable wedge. The trigger to go all-in came from realizing the incumbent's core logic was bad — not its design. Familiar UX plus a genuinely better engine is enough to displace.
“The tip number one for sure will be to validate before spending money on ads. And when I say validate, it's not just that the product works, but also that people are willing to pay for it.”
Validate willingness to pay before you spend a dollar on ads
Free signups and engagement are not proof of a viable subscription business. Validation has to include actual paying customers before any paid acquisition turns on — otherwise the ads are funding a leaky bucket disguised as growth.
“Initial MVP took around two to three months maybe… I use Cursor a lot, but I'm very specific which files to touch and don't let the AI just go rogue basically.”
Ship the MVP in 2-3 months and keep the AI on a tight leash — specific files only
The first version shipped in two to three months by scoping tightly around a tracker plus a custom workout engine. AI coding tools accelerated the work, but only when pointed at specific files rather than allowed to roam — autonomy is what produces unmaintainable sprawl.
“I posted this thread on how I built Gravl… I shared kind of the technical specs around it. It got over a couple hundred likes within the first couple hours and over 300,000 impressions. Yeah, we got our first couple thousand users.”
A single Reddit build-story thread can net 300K impressions and your first thousand users
The first thousands of users came from a single Reddit thread describing how the app was built, aimed at developers who happened to also lift. Sharing the technical build story (not a sales pitch) earned 300K impressions and the feedback loop that confirmed product-market fit.
“UGC is king. You'll see that most of the ads that work out there are UGC content. You'll see a lot of AI videos — the cost of producing them is cheap, and they're easy to test.”
UGC is king in consumer ads — pay $50 a creator and lean on AI video to multiply volume
For consumer apps, user-generated-style content beats polished production in paid ads. Pay creators as little as $50 per piece, lean on AI video tools and CapCut to multiply volume, and treat creative as a testing problem rather than a craft problem.
“Good thing about Meta Ads library is that it's all public. You can go into any of your competition's dashboard and see which ads are working for them, where they're putting more money… Just copy what works for them and start like that.”
Copy winning creatives from your competitors' Meta Ads Library — it's all public
Competitor ad libraries are a free shortcut to a working creative strategy. Watching which ads competitors keep running reveals what is converting, so new entrants can model proven hooks and formats instead of starting from scratch.
“We get your name and then we'll just go and ask you a series of questions — like why you want to use a fitness app, we ask you about your experience.”
Onboarding doubles as a personalization quiz — every answer feeds the AI engine
The first-run flow is a guided questionnaire covering goals, experience level, training frequency, and equipment. Each answer feeds the AI workout engine, so onboarding itself becomes the value proposition that justifies the paywall waiting at the end.
“Hard paywall before you sign in for the first time. It's like, how many users are actually going to pay before they actually see their product? And the answer is a lot.”
A hard paywall before the user ever sees the product — a lot of them still pay
A hard paywall placed before any product access converts far more users than intuition suggests. Gating the experience at the moment of peak intent — right after a personalised onboarding quiz — is what makes a consumer subscription app's economics work.
“We have a 24/7 support chat inside the app, and that includes everything from articles to messages. We'll have someone reply — an actual person, not an AI. That's something that users value a lot.”
24/7 human in-app support — not bots — is a real retention lever in subscription apps
In-app support staffed by real humans, not bots, is a major growth and retention lever in consumer subscription apps. Users notice and value the difference, which compounds into long-term loyalty and word-of-mouth referrals.
“We translated the app to Spanish, we started running ads on South America, and we were spending less than 50 bucks a day on ads. And that's what worked for us from the start up until this day.”
Translate to Spanish and run cheap South America ads — $50/day buys real volume
Translating the app and running ads in South America let the team scale on under $50/day in CPMs. Skipping the expensive US auction at the start and using a native-language advantage is how the paid engine was built — geo-arbitrage on Meta and TikTok still works.
“Our expense number one is Meta and TikTok, then maybe a little bit of Google, maybe a little bit of Apple search. I'd say a third is somehow accurate — a third of our revenue, that is.”
Paid ads cost roughly one third of revenue — track CAC as a share, not a number
Paid acquisition runs at roughly one third of revenue, with Apple's 15% cut and a small infrastructure plus AI bill on top. A bootstrapped consumer app can stay highly profitable when CAC is tracked as a fixed share of top-line rather than an absolute number.
“In our case, it was our previous startup — we probably dragged that for an extra year. We should have probably called it days like way earlier.”
Knowing when to kill the previous bet is as important as persistence
Persistence is the default advice, but dragging a dying startup for an extra year is a common and expensive mistake. The discipline to call time on a non-working product is as important as the discipline to keep going on one that does.