0 —> $10k day in 41 days.
Forgot to post it but cool seeing it. Very grateful.
Hedonic treadmill is real, $100k day soon.
arqo retweeted
I just started using grokbots today and holy shit.
EVERYTHING is in one central dashboard.
Jarvis is the master that monitors my "units" which includes a team leader/media buyer, copy chief, video/image editor (Kie.ai) and ad research (still need to add gethhookd API).
I have a campaign scheduled to go live at midnight on Meta, let's see what happens.
Also HUGE shoutout to @KrsteKovaceski for making the playbook 💯
arqo retweeted
If an objective does not violate the laws of Physics then the only thing stopping you from reaching it is the knowledge in between.
arqo retweeted
Secret of the DTC Universe #2: When it comes to ad copy - 90% of your effort should go into the concept and hook.
This is true for basically any type of Meta ad at least:
-A 30 second side-by-side comparison video with no voiceover at all…
-A two minute long yapper ad…
-A 3,000 word native style Meta caption ad.
Virtually ALL of your work should be on finding the right concept and hook.
Why?
Because the rest of the ad is just a formulaic outline that can be swiped.
This is especially true with LLMs.
The process is so easy, and there are just 4 Steps.
Step 1. Give the LLM an ad from Brand A and ask it to analyze the structure, voice, and word count…
Step 2. Give it a new ad concept or hook for a different brand (Brand B).
Step 3. Include some basic context for Brand B - either in the chat, or as part of a project.
Step 4. Tell the LLM to write a new long-form native ad for Brand B that models the structure and word count of the ad for Brand A.
Voila!
99% of the time you get a good ad.
That part takes like 1-2 minutes.
Maybe another 10 minutes for small tweaks to the first draft of the ad.
It’s all super easy.
The part that’s not so easy though…
Is the work that goes into getting killer concepts and hooks.
And so then, of course, the question becomes:
“What makes a great hook?”
Well, let’s talk about that next.
A great hook is curiosity driven, emotional, or both.
It is RARELY about the product.
It OFTEN starts in the middle of the action.
It NEVER tries to open and then close a loop in the same breath.
And it ROUTINELY comes from outside the ad.
Sounds good, but let’s look at this in practice, shall we?
Let’s say you’re doing a 3,000 word long-form native style Meta ad.
For those who are uninitiated…
These are ads on Meta where you have an organic-seeming image without text or captions…
Then the caption copy is this super long “mini sales letter” driven by an emotional story from the spokesperson…
That ultimately leads to the discovery of whatever product the brand sells.
So for example, let’s say that the brand sells a no-pull dog harness.
The native ad might be written from the perspective of the adult daughter of an elderly woman who used to love walking her dog, and then stopped.
The ad would set up that the spokeswoman’s mom really loved that dog…it was her best friend…and she loved their walks…
But then, suddenly, the walks slowed down and stopped.
At first the mom plays it off…
But eventually she confesses to the daughter that it’s because she doesn’t feel strong enough to control the dog anymore when he pulls.
Her shoulders ache, her back aches, she worries about the dog running off.
Ultimately, her deepest pain point is feeling old, and weak, and like she’s lost her agency and control.
Then someone (the elderly woman, her adult daughter, a friend) discovers the no-pull dog harness.
The mom tries it out, she’s nervous, it works. She goes back to walking her dog and is happy.
And of course there’s some kind of call to action.
Broad sketch, but you get it…
And here is an example of what would NOT make a good hook:
“My 77 year-old mother used to love walking her dog. Then she stopped. Not because she didn’t want to. But because she was afraid she’d lose control. Here’s how a no-pull harness gave her back those precious mornings.”
I wrote this manually, on the fly, without AI…
But it’s fairly representative of the kind of AI riff-raff you get when you ask for hooks.
And here’s what’s wrong with it:
*First - it tries to close the whole loop.
This is a really common AI issue.
Most LLMs want to “close the loop” and resolve things…
Whereas great direct response marketing opens loops and keeps them open for an uncomfortably long time.
This goes back to curiosity…
Creating a great hook is akin to creating an itch that your prospect can’t quite scratch.
They want to scratch it. They need to scratch it. They can’t stop thinking about how good it will feel once they do scratch it.
But they can’t reach that itch right away…
They have to wait to get relief.
*Second, and related to the first point…
What the LLM is really doing is trying to tell you the entire story of the ad while pretending that’s a hook.
This is another way of saying the same thing I did in point #1…
But it’s worth calling out…
Because it’s a super common pattern with LLMs.
Look at the hook again. It’s literally just telling you what happened in the ad.
There’s no mystery.
There’s no curiosity.
There’s no real reason to keep reading.
*Third, it talks about the product when, for this type of ad, it shouldn’t.
The whole point of a long-form native style Meta ad is that it feels like this organic post. The fact that the “poster” discovers and then recommends a product should feel almost accidental. It just so happens that this product was the best solution, and they want to tell you about it.
–
Okay, so what are some examples of much better hooks/openings?
Let me riff:
Hook Option 1:
“My mom walked that dog every morning for 11 years. And then, one Tuesday, she just stopped.”
Hook Option 2:
“My mom’s the strongest woman I know. And she quit on that dog.”
Hook Option 3:
“The dog was laying there by the front door. It broke my heart.”
Hook Option 4:
“I saw Freckles laying by the front door. It broke my heart.”
Hook Option 5:
“At first, I was mad at my mom. Then, I felt sorry for her.”
Hook Option 6:
“I called the rescue shelter on a Tuesday. My mom found out the next morning, and I thought she’d never speak to me again.”
–
Okay so those are just examples…
But let’s highlight why they’re better than the initial example hook.
The biggest reason is this:
They are all open loops.
They convey emotion.
And they create curiosity or conflict, but don’t resolve it.
They also often start in the action:
“Was laying” // “I saw” // “was mad/felt sorry” // “she quit” // “she stopped.”
And while there’s no actual ad to reference…
Some of them could have come from outside of the ad too.
For example - maybe there’s nothing in the ad about the adult daughter calling a rescue shelter.
That’s fine. You don’t really need to revise the ad.
It can still open with the idea that she called, her mom found out and got mad...
And then the adult daughter just goes “let me explain” and it transitions into the rest of the ad.
—
Okay, this is a long one.
Hopefully it’s valuable though…
And if it is, let me know…
Because I have dozens of additional secrets to share…
But I’m only going to keep going if people keep reading and engaging.
arqo retweeted
jumped on a call with this mf he's legit af and me and him about to do 1mil/d with our infinitaley scalable agentic editing infrastructure that learns from very well edited ads. i want to give a polite thank you to every hard working CS and Editor for fueling the data points of my ripping empire. Kisses. Bye!
few insights on our creative process to hit $200k day in less than 3 months:
1. traditional creative strategists are obsolete, our creative team is literally 3 editors (@argon_delta)
2. manual intent no longer exists (sort of)
with the tools that exist, we were able to step away and play 2k and geometry dash..
while all research, swiping, scripting, editing, and learning were fully automated
there is no need to waste time reading breakthrough advertising or manually trying to learn and study your customers thinking patterns
the focus for us was on optimizing and automating our entire workflow to pump out ads that are as good as whatever the top ecom brands are doing right now + focusing more on volume rather than intent to learn much faster
3. there is zero need to reinvent the wheel, swipe what works and remake for your own product.
"learning" how to make ads doesn't exist anymore
whatever people say about "you can't just oneshot ai to take a swipe and make it for urself" is false to an extent.
your ego is stopping you from handing off creatives to ai because you keep reading that it isn't good enough yet.
3. EVERYTHING is automatable with ai
there is nothing in the process of making creatives that requires human work anymore, and if you don't get ahead you will get gapped very shortly during this new meta of ecom.
may or may not talk more about this in the future
there is a LOT of gaps i'm not filling in this tweet, if you're pushing some solid rev feel free to dm me about anything i mentioned.
arqo retweeted
i've been saying this for awhile
October is often a softer month than September
focus on efficiency...
you'll see the same aggressive scale as last year in November and December
Something is seriously off with Meta traffic right now.
We're seeing conversion rates drop by as much as 30–40% across parts of our portfolio over the last 3 weeks.
And it's not just e-commerce.
DTC, B2B, lead gen. Different industries. Same pattern.
Here's the weird part:
CPMs? Fine.
CTRs? Fine.
Add to carts? Fine.
Initiated checkouts? Fine.
Actual conversions? Falling off a cliff.
The brands struggling the most have one thing in common:
They're NOT running aggressive promotions.
Meanwhile, brands that have been pushing offers through Labor Day, Prime Day, and now Halloween?
They're doing just fine.
We've seen similar behavior around major shopping events before, but I don't remember seeing it at this scale or for this long.
And after speaking with other Tier 1 advertisers, I'm hearing very similar stories.
Which makes me think something bigger is happening.
Two possibilities:
1. Meta is making significant backend changes, as they frequently do, with very little communication about how those changes affect delivery and traffic quality.
2. We're seeing broader macroeconomic or consumer behavior shifts that are starting to affect even Meta's ability to find high-intent buyers.
And that second possibility is particularly interesting.
Meta has historically been exceptionally good at finding people who are ready to buy, even when the broader market isn't doing particularly well.
But right now, something isn't adding up.
I don't know which explanation is right. Could be a combination of both.
What I do know is that this isn't isolated to a few underperforming accounts anymore.
What are you guys seeing?
arqo retweeted
ill drop some sauce too.
when generating with seedance 2.5 480p, make sure to add a 4k reference photo. it will make the quality really high
Quick tip to save huge amount of $$$ on credits for AI Drama and AI UGC:
Use kie.AI instead of higgsfield
You can try seedance 2.5, will save around 2/3 the cost. What would cost $300 in credits in higgs, we make with $150
For even further cost savings, you can use WAN to cut cost to around 1/5th
arqo retweeted
Guys come into dropshipping and fall in love with a new product or store right away
Literally have the game backwards
They saw the guys with brands and thought they had to that too
End up doing all this overthinking and other bs and start wasting energy and time
This isn’t the typical business model my G
Run ads first
If it hits cool, now go setup the backend
If not, keep it pushing
You don’t need warehouses, bulk, or a branded product
You just need a day or two of your full focus
And if it doesn’t work, basically keep trying, but don’t be dumb, take learnings and improve every time
This is the hardest part though… that feeling after a failed launch hits different… it puts a lot of mfs into permanent gulag
Meta being Meta again
Dont kill your winners because the account looks weird for 2 days
Thats exactly how people kill their best ads before BFCM
arqo retweeted
I would’ve made a lot more a lot quicker if I didn’t exclusively try to sell supplements for the first year dropshipping
few insights on our creative process to hit $200k day in less than 3 months:
1. traditional creative strategists are obsolete, our creative team is literally 3 editors (@argon_delta)
2. manual intent no longer exists (sort of)
with the tools that exist, we were able to step away and play 2k and geometry dash..
while all research, swiping, scripting, editing, and learning were fully automated
there is no need to waste time reading breakthrough advertising or manually trying to learn and study your customers thinking patterns
the focus for us was on optimizing and automating our entire workflow to pump out ads that are as good as whatever the top ecom brands are doing right now + focusing more on volume rather than intent to learn much faster
3. there is zero need to reinvent the wheel, swipe what works and remake for your own product.
"learning" how to make ads doesn't exist anymore
whatever people say about "you can't just oneshot ai to take a swipe and make it for urself" is false to an extent.
your ego is stopping you from handing off creatives to ai because you keep reading that it isn't good enough yet.
3. EVERYTHING is automatable with ai
there is nothing in the process of making creatives that requires human work anymore, and if you don't get ahead you will get gapped very shortly during this new meta of ecom.
may or may not talk more about this in the future
there is a LOT of gaps i'm not filling in this tweet, if you're pushing some solid rev feel free to dm me about anything i mentioned.
arqo retweeted
Saw a guy call it cuck coding instead of vibe coding because you just sit in a chair and watch😭
Meta going through and rejecting nearly every single ad in my ad account as we speak, all were already running for multiple days some even weeks.
Tried to replace some with dog photos and even some of those were getting rejected😭
Yea my shit getting nuked tn. What a start to Q4 man.






























