Three steps on the right. Only one of them is hard.
Everybody has forty ads now. Almost nobody has one that works.
The hard part was never the camera. It was knowing what to point it at and that part didn't get automated. It got buried under a hundred accounts insisting it did.
Which is why the right-hand column above ends with "cut" and not with "profit." Generating is solved. Writing the thing worth generating is not.
We teach the part that makes your ads better.
Now it takes about 40 minutes and roughly $3 per video. The bottleneck was never ideas. It was the time and cost required to make enough creative.
When one video takes weeks and costs hundreds, volume is the constraint. When it takes minutes and costs a few dollars, testing becomes the advantage.
A brand owner doing $700k a month and a freelancer on his fourth client have the same problem. Both belong here. What matters is whether you already ship.
Every prompt we use, tested, sorted by format — talking head, product demo, animation, statics, b-roll. Start from a known-good result and adapt it. Never from a blank box.
Step by step, in the order you do them, written down. Hand them to an editor or a VA and the work happens without you.
We take apart ads winning right now — ours and yours — and show the reasoning, not the output. Bring your work and get it pulled apart.
Step one of the right-hand column, and the only one that decides anything. How to read a winning ad, find the angle underneath, and write to it.
What we subscribe to, what it costs, what to cancel. Most people pay for two tools that do one job. This section often costs less than nothing.
People who ship, not people asking how to start. Somebody solved your Tuesday problem last Tuesday.
A freelancer in our survey said what he actually wanted better than we could: "exact frameworks, not just very surface level steps."
That's the gap. Free content shows you the result and skips the reasoning — which model, which settings, which eleven things went wrong first.
It's built to look impressive in sixty seconds. Not to work on your product on a Tuesday.
A prompt without the reasoning behind it is a screenshot of somebody else's thinking.
Two ads for the same product, made the same week, generated the same way.
Same generation settings. Same production cost. Same afternoon.
The difference was the sentence somebody wrote before any of it started.
That sentence is the whole job. It's also the thing every operator we surveyed said AI is worst at.
The SOP is open in a second window and you check every step before you do it. You're slower than last week. You also stop making mistakes, and you finally know why the output looks how it looks.
The SOP is closed. Video and statics, more of both, on a trend the same week it appears. The library has your own prompts in it now.
Deliberately unimpressive. Anyone promising a transformation by the weekend is selling you the demo, not the job.
In case you think this is only our opinion.
180 of the 400 operators already make a video for under $300. 96 have already tried AI ads and 48 had them perform. Cost stopped being the constraint a while ago — here's what replaced it.
Anonymised 0xROAS market survey. Respondents could name more than one frustration, so mentions exceed the sample.
The brand owner doing $700k a month and the freelancer on his fourth ad gave us the same answer. Revenue doesn't predict this problem. Having a system does.
Somebody wrote almost exactly that in our survey, and they were right to be annoyed — it's the most common way these things waste your money. So the live calls aren't lecture repeats. They're teardowns of ads running this week, including yours, with the reasoning shown. If a call could have been a recording, we shouldn't have held it.
Nothing about the tools. What's new is step one of the right-hand column — the writing — plus a library so you never start blank and SOPs so it survives you being busy.
Second most common answer in the survey. The cause is nearly always the same — you're describing what you want instead of controlling what decides it. Specific, teachable, and where we'd start you.
Four respondents said exactly that. It's why the stack list exists, with real prices and an honest note on what to cancel. A fair number of members leave that section spending less per month than before they joined.
No, because we turn beginners away. If you've never run an ad the free material is the right start, and we'll tell you so. The room is only worth $148 if everyone in it ships.
"Insane game inside. Anyone who isn't here now is going to hit themselves in three months. I'm getting ads made for $3.15 each in 15 minutes and they're performing just as well as humans."
"Not only how to create hyper-realistic AI UGC, but how to create winning ads by analysing already-winning ads. It's not an AI UGC course, it's a complete ad domination system."
"Everything with AI seems so overwhelming with new tools coming out almost every week — it's easy to get lost in the sauce. 0xROAS really comes in clutch to help marketers get up to speed."
One month costs less than one afternoon of failed attempts.