Ross set himself a challenge: make $1 with AI in less than 24 hours.
It took him over 30 days.
In between: a month of excuses, a launch tweet that got called "AI slop" by every single commenter, an app he pulled down the same day he shipped it, and one very long night of wanting to quit.
This is what a failed product launch actually looks like from the inside. Not the sanitized version founders post after they've recovered. The raw one, filmed while it was happening.
Ross runs the $100K Build Challenge on Starter Story Build: make $100,000 using only AI tools, document everything, no quitting allowed. Day 9 is the episode where it all went wrong, and it's the most useful one so far.
Watch it here, then I'll break down what I'd steal:
The app that took a day to build and a month to launch
The idea came from a pattern I see constantly: find something already making money, apply it to a different audience.
Ross had found an AI resume generator doing $20,000 a month. Simple app, obvious value, easy build. So he asked the right question: what's the same mechanic for a different crowd?
His answer was Build in Public. You fill in an onboarding flow (what you're building, what you're struggling with, who you are) and it generates a content strategy plus your first 30 build-in-public tweets. He even bought the domain, gobuildinpublic.com.
Cursor built roughly 90% of it in a single day.
Read that again. One day from idea to near-finished product.
Then it sat unlaunched for a month.
The excuse zone
Ross has a name for what happened next, and I think every founder should adopt it: the excuse zone.
The app was done. So his brain started manufacturing reasons not to ship it:
→ Week 1: his wife was graduating residency and family was visiting. "If I launch this and it takes off, I won't have time for support."
→ Week 2: moving across the country, Florida to California. Next week.
→ Week 3: a planned trip to Europe. Surely you can't launch before a vacation.
→ Week 4: "I'll launch when I'm really ready."
A month gone. Zero launches.
Here's the part that stuck with me. Looking back, Ross admits the excuses weren't really about time. Deep down he knew something was wrong with the product, and the excuses were a socially acceptable way to avoid finding out.
I've done this. You've probably done this. The side project that's "90% done" for six months isn't a time problem. It's a fear problem wearing a time problem's clothes.
Launch day: "great, more AI slop"
Ross finally shipped. One tweet, a link, a short demo video. Nothing fancy, just trying to get early users and feedback.
The tweet barely moved. A couple of views. Then the first comment landed:
"Oh great. More AI slop in the world."
Then another: "This is despicable. Now I get to see more AI garbage in my feed."
Ross is used to rough comments. He's on camera every week for one of the biggest founder channels on YouTube. But this was different: every single comment was negative. Not one person defended the product. Not one person was curious.
He took the app down the same day. Threw in the towel, demoralized, done.
And this is where the challenge rules saved him. Rule three: you're not allowed to quit. His team at Starter Story told him the only sane thing you can tell a founder on launch day: shut it down, sleep on it, look again tomorrow.
(If you've never had a launch die in public, it's brutal. I've had launches do 30k views and launches do crickets, and the crickets hurt more than the insults.)
The overnight repositioning
The next morning Ross did something most founders never manage: he agreed with his critics.
Not completely. But enough. The product was rushed. It wasn't something he truly believed in yet. And the positioning invited the exact attack it got: a tool that generates 30 AI tweets for you is, by definition, an AI slop machine. The commenters weren't wrong about the product. They were early.
But the problem underneath was still real: builders who ship great stuff every day and are terrible at telling anyone about it.
So he kept the conviction and killed the positioning. Instead of a done-for-you tweet generator, he rebuilt it as a tool. An AI coding tool, but for marketing. You give it context (what you shipped today, what problem it solves, who it's for) and it drafts content in formats trained on posts that actually performed. It asks you questions the way Cursor asks about your codebase.
He rebuilt the whole thing with Claude Code, which was blowing up right around that time. New UI, new framing, new name: MoAT.
The name is the thesis. As AI makes building trivial, distribution may be the final moat. Your audience, your brand, your building in public. That's the one thing a competitor can't clone in a weekend with the same tools you used.
Same tech. Same underlying problem. Completely different product. That's what one honest night of reflection bought him.
The micro launch that finally made $1
For launch number two, Ross did the opposite of launch number one.
No launch tweet. No public announcement. He picked two people who badly needed the product and got them using it. That's the whole strategy: make it work for one person, then ten, then a hundred. He calls it a micro launch, and he's building a waitlist behind it while the beta runs.
Early proof it's working: one user roughly doubled their followers off a tweet partly generated in MoAT. And one of those beta users converted into a paying customer.
$1. Over 30 days late. Still counts.
Ross is upfront about the math: "I'd be better off driving for Uber or doing Instacart orders." He's right, for now. But a wage doesn't compound. One stranger paying for something you made is proof of demand, and demand is the thing you can grow.
His channel-mate Alex went through the same emotional arc across 70+ shipped apps and $0 in revenue before fixing his approach. I broke that one down in the Starter Story Challenge post. Different builder, same lesson: keep shipping, and be honest about what's actually broken.
What I'd steal from Ross's failed launch
Five things worth taking from this episode:
→ Name your excuse zone. If a finished product has waited more than a week, you're not busy, you're avoiding a verdict. Ship it or admit what you're afraid of.
→ A failed product launch is data, not a verdict on you. Every comment on Ross's launch was negative, and every comment pointed at the same fixable thing: positioning. Angry feedback is still feedback.
→ Sleep before you decide anything. The only decision Ross made on launch day was to stop making decisions. The repositioning idea showed up the next morning.
→ Reposition before you rebuild. The tech barely changed between the flop and MoAT. The framing changed everything. Ask what your product is a "coding tool for marketing" version of.
→ Launch small when trust is low. Two beta users who desperately need you beat a thousand strangers who don't. If your product touches AI content, the internet's default reaction is hostility. Earn receipts privately, then go loud.
And the meta-lesson: Ross only survived this because quitting wasn't an option he'd left himself. Build your own version of rule three. Mine is building in public: once people are watching, disappearing costs more than continuing.
FAQ
What should you do after a failed product launch?
Nothing, for 24 hours. Ross took the app down, slept, and looked again the next day with distance. Then separate the signal from the sting: his commenters attacked the positioning ("AI slop"), not the underlying problem. He kept the problem, rebuilt the framing, and relaunched quietly to two beta users. Most failed launches are positioning failures, not product failures.
How long does it take to make your first dollar online?
Longer than the YouTube thumbnails promise. Ross set out to make $1 with AI in 24 hours and it took him over 30 days, including a full rebuild. Most founders I've covered on this blog took 4 to 6 weeks just to ship version one, before any revenue. The honest range for a first dollar is weeks to months, and the excuse zone (not the build) is usually what eats most of that time.
Should you relaunch a failed product or kill it?
Ask whether the problem is still real. Ross had proof that builders struggle to market what they ship, so the problem survived even though the product flopped. If conviction on the problem is intact, reposition and relaunch small. If you only had conviction in the solution, kill it. His Build in Public app died, but MoAT was built from its parts in days.
What is a micro launch?
Launching to a handful of hand-picked users instead of the public. Ross picked two people who badly needed MoAT, worked with them directly, and let their results drive the next iteration while a waitlist builds in the background. It trades launch-day dopamine for real feedback and testimonials, and it protects a fragile product from the "AI slop" pile-on a public launch invites.
Steal playbooks like this every week
Ross documented his worst launch on camera so you don't have to learn this with your own product.
Every week on the Profitable Founder Podcast, I sit down with bootstrapped founders doing $100K to $10M a year and pull apart exactly how they did it. The failures included, because that's where the playbooks hide.
