how to spot a grifter on ecom twitter:
they larp having 2 ROAS or 3 ROAS but they never show how much the ad have spent (i wonder why)
2 ROAS on $1,000 spend is not a flex.
1.5 ROAS on ad with $250k spend is.
once your ads start spending high hundreds of thousands a month, they'll start showing up on atria, trendtrack, twitter, etc... and EVERYONE will start copying you
so if you keep running the exact same shit over and over again, after a few weeks your ads will start looking like every other ad in your niche.
keep doing what works, but keep evolving it. new formats, new unique stories, new angles, new iterations, etc...
if you stop moving the needle, your copycats will literally outgrow you
30 REASONS YOUR ADS ARE SHIT
(from an 8 figure creative strategist)
1. the hook sounds like an ad right away
2. you copied a trending format, not a winning skeleton
3. your ad avatar talks way smarter than your customer
4. your ad blames the fault on your customer
5. you didn't read the reviews and apply learnings
6. you didn't read the comments and apply learnings
7. you're not testing enough ads
8. you don't iterate your winners enough
9. your unique mechanism is not unique
10. you use complicated words
11. the offer is the same as your competitors
12. you talk about your product, not about how it's gonna solve your customer's problems
13. there's no reason to buy today (no urgency)
14. there's no authority (trust problem)
15. your script sounds obviously like ai
16. you're not giving feedback to your creative strategist
17. the product shows up too early (the ad feels too salesy)
18. the ad is nicely edited but says nothing valuable
19. you're copying your competitor's ads 1:1
20. your angles are the same as 10 other brands
21. your ad is terribly edited (execution problem)
22. the claims sound made up and not believable
23. you let ai do everything and didn't work WITH it
24. the script was written for the brand owner's ego, not the customer
25. you only make ads you personally like (not what your customers like)
26. the cta is weak or missing
27. you make just 1 version of each idea
28. you let your editors write the scripts
29. you do everything manually
30. you didn't read breakthrough advertising
I'M BEGGING YOU DON'T PLAY GTA VI TODAY
> instead, build a team of ai agents in grok bot
> give every agent one job
> let them swipe competitor ads
> iterate the scripts on a call with ur copywriter ai agent
> make 100+ ads a week
> get rich and play GTA irl
here's exactly how to do it:
1. CREATE A ''MANAGER AI AGENT''
one AI agent that runs the whole team. u only talk to him, he talks to the rest of the ai agents.
without a manager u end up with 10 chats open and no idea what's going on
2. GIVE EVERY AI AGENT ONE JOB
> ads scout
> researcher
> copywriter
> QA
that's enough to start
one agent per job. why? cuz an agent that does everything is average at everything..
every agent gets its own prompt and rules (what to do, what NEVER to do, what good examples look like,etc...)
3. MAKE IT RUN WITHOUT YOU
set routines so they work every day on their own
scout finds new competitor ads, pulls the transcripts and turns them into ready to send briefs (with the original script inside)
you wake up and the briefs are already done
4. ITERATE ON A CALL
jump on a call with ur copywriter AI agent and write the script together
adapt it to ur product and niche, make it better, add 2-3 new hooks, etc...
it's like writing a script with a junior copywriter while looking over his shoulder. explain every change and WHY you're doing it
you'll get a great script, your copywriter ai agent remembers it and gets better next time, and u learn the skill better yourself as well
(a 2011 study found students who taught material improved their test scores by 23 percentage points vs 11 for material they only prepared)
5. QA BEFORE ANYTHING LEAVES
one AI agent checks every script against your rules before it gets sent to your editor
wrong claim, boring hook, too smart for ur customer = back to the writer
6. THE MATHS
i run 50+ agents rn that helped me 10x my output volume as a creative strategist for 8/9 fig dtc brands.
i put together a pdf with this exact setup. every agent, every prompt, every routine
comment "ADS" and follow me so i can send it over to you
no skill can make you rich faster than being creative strategist for a few dtc brands
you can go from dead broke to making $30k-40k/m profit in 6 months
no product, no inventory, no bs, just grinding new ads every single day
if the guy/girl talking in your ad sounds way smarter than your average customer, they're not gonna trust you
when the customer doesn't understand something in your ad, it doesn't feel impressive, it feels uncomfortable and unfamiliar
and uncomfortable = they're gonna scroll (back to comfortable)
this is where most creative strategists fuck up, they want to sound smart and use fancy words, complicated mechanisms like ''clinically proven bioavailability'', etc...
SPEAK LIKE YOUR CUSTOMER MF
if your ads tell your customers they've been doing something wrong the whole time, they're not gonna buy
buying from that ad would mean admitting they're stupid
you always have to blame everything else except them - the other products, categories, companies, influencers, scammy ads, "evil" corporations, etc...
but NEVER blame your customer
make them feel like it's not their fault. those other products were just never made for their specific problem
ai agents and agentic workflows will never turn a bad creative strategist into a good one
they just make him produce 10x more shitty ads, faster
they're amplifiers of a creative strategist's taste and judgment (which can only be developed through experience and reps)
if one of your ads has been top spender in the ad acc for 2+ days straight after launch, and you are working on ''new ideas/concepts'' instead of absolutely fucking squeezing the winner ad in 100 different ways, you're doing it wrong
6
36
1,610
''ai ugc doesn't convert, people can tell it's fake''
skill issue
i've written dozens of ai ugc ads that spent high 6 figures (each)
a real creator saying an average script will always lose to an ai avatar saying a great script
8
20
1,244
imagine still paying $1k/month for Indian or Filipino VAs in 2026 😭
I fired my entire VA team and replaced them with teams of AI agents that work together just like human VAs
but they're smarter, faster, cheaper, and they work almost 24/7
and they never come up with stupid excuses like: "hello sir, we have a power outage for the next 2 days, so I can't work."
5
40
4,289
if you're not building custom internal tools and apps with ai agents, you're ngmi, and missing out on sooooo much money lol
if you're using agentic workflows (ai agents), you know that 100+ ads a week is not hard
production isn't the bottleneck anymore
but if most of those 100+ ads are the same ads, just with a new avatar, a new hook or new visuals, then your volume will go UP, but your hit rate will PLUMMET
all that speed is only worth it if someone with experience, skills, knowledge and taste decides what the ai agents are going to be working on
that's why ai agents aren't replacing creative strategists. they make them more important than ever
''Ai SoNg aDs'' 😭
''DrAmA aDs'' 💀
''cLaYmAtiOn aDs'' 🥀
everyone's copying the format and ignoring what the ad actually communicates. the ad format is NEVER the reason someone buys your product
brands like Resilia aren't winning because their ads are animated songs lol. they're winning because of what the ad communicates. it calls out the exact problem, destroys the viewer's doubts, and shows a transformation that feels realistic to them (they believe they can achieve the desired result as well)
put that same messaging in a well executed ai ugc and it would probably still work
so when you see an ai song winner, don't just copy the song. break down the exact skeleton:
1) what it communicates
2) why people trust it and buy from it
3) all the winning elements
4) then apply those elements onto your product and test that skeleton in all formats that can utilize storytelling (songs, dramas, yappers, animations)






