
Deep in the ecom trenches since before it was cool. Built @pinnaclemediads Now building @pinnoradotai & scaling things that weren’t meant to scale.
Dubai, United Arab Emirates
Joined March 2013
- Tweets106
- Following502
- Followers281
- Likes352
All Tools > Customer experience Insights (On your BM)
Meta is quietly giving your brand a credit score
and most of you have no idea yours is sitting in the bottom 5%
it's called the customer experience score. meta asks people who bought from your ads how it went, then ranks you against every other advertiser
low score = your ads reach fewer people for the same budget
so your creative's fine. targeting's fine. CPMs still going up because:
→ order showed up 2 weeks late
→ product looked better in the ad
→ refund took forever
→ support left them on read
you can't fix bad ops with better ads anymore
1
291
Meta is quietly giving your brand a credit score
and most of you have no idea yours is sitting in the bottom 5%
it's called the customer experience score. meta asks people who bought from your ads how it went, then ranks you against every other advertiser
low score = your ads reach fewer people for the same budget
so your creative's fine. targeting's fine. CPMs still going up because:
→ order showed up 2 weeks late
→ product looked better in the ad
→ refund took forever
→ support left them on read
you can't fix bad ops with better ads anymore
Swarith retweeted
I’ve managed $500M+ in ad spend across 200+ brands over 20 years. Some of them my own.
For Dr. Axe:
Before: $4.1M spent, $15.2M back.
After: $12.9M spent, $55.7M back.
3x the spend. And the return went up, not down.
Most brands watch ROAS fall apart when they scale. Ours climbed from 3.67 to 4.30.
The budget wasn’t the secret. The thinking was.
And most of what goes viral on X about “scaling” right now gets that thinking backwards.
Here’s what half a billion dollars taught me that nobody’s posting about:
A format is not a strategy
Every few months the timeline falls in love with a format.
Pixar ads. Song ads. Fake news ads. Now drama ads.
Brands copy the format, then wonder why it flops for them.
Here are 3 of the Dr. Axe ads:
A crying selfie in a car. $1,967 spent, $47,938 back.
A woman on her bathroom floor. $6,003 spent, $69,488 back.
A plain green box with text on it. $2,432 spent, $44,836 back.
Completely different looks. Same thing underneath: they said what the customer was already thinking.
The format is the wrapper. The angle is what sells.
Your customers already wrote your best ads
The best hooks I’ve run lately came almost word for word from Reddit.
Women kept posting the same thing:
Their cabinets were full of supplements. They didn’t know what was working. And they were still exhausted.
So we stopped selling another bottle. We sold the way out of the cabinet.
The ads built on that one idea: $59K spent, $351K back.
A simple before and after from that angle turned $14K into $107K.
Across the 10 ads below: $117K spent, $696K back. Almost 6x.
And these 10 are a small sample of what ran inside that $12.9M.
Read your reviews. Read your competitors’ reviews. Read Reddit.
Your customers are telling you exactly what to say. Most brands never listen.
Volume without research is expensive guessing
Everyone says “make more creative.” True.
But 500 ads on the wrong angle is just 500 ways to lose money faster.
Here’s one of my own brands last New Year’s:
Dec 30: $802
Dec 31: $4,842
Jan 1: $38,177
One creative change.
Not a new product. Not a sale. Not influencer buzz.
One ad, built from research: who’s buying, what they’re saying, what’s stopping them. Then we let it run.
That’s when creative and algorithm line up.
Every winner has an expiry date
The ad printing money today is already wearing out.
The fix isn’t panic switching formats. It’s having your next angle ready from the research before you need it.
The brands that scale don’t wait for a winner to die. They’re already testing its replacement.
Study the customer, not the platform
Algorithms change every quarter. People don’t.
I’ve lived through every Meta update that was supposed to change everything. iOS 14 included.
The fundamentals never moved: the right message, to the right person, at the right moment.
If you’re trying to scale this Q4:
Stop copying formats. Find the angle first.
Mine your reviews and Reddit before you brief a single ad.
Build your next winner before your current one dies.
I’ve spent 20 years doing this behind the scenes. Time to start sharing it here.
Follow along. Lots more coming.
Thousands of pet brands are running the same “cute dog + product shot” ad
the easiest way to stand out: take a format people already love and point it at something they see every day
so i gave a golden retriever the full nature documentary treatment
calm narrator. serious tone. completely normal dog behaviour
“every blade of grass is a news report. most of it, frankly, is about the neighbour’s cat.”
he rejects the rug, briefly considers the floorboards, then finds the cooling mat
“the search, it seems, is over.”
the product shows up as the discovery, not the pitch. and every pet owner watching sees their own dog
no camera crew. no dog trainer. no waiting for the perfect shot. 100% AI
the narrator, the dog, the script, every shot. in under 2 mins. all with pinnora.ai 🐕
Cease & Desistfield
Last night, @higgsfield sent @ComfyUI a cease and desist (attached)
They want the Open Source Higgsfield challenge taken down within 48 hours.
Their argument is that our use of the Higgsfield name and that using their CEO's likeness is an infringement violation (man, the irony). Claiming that it makes the event look like Higgsfield is associated with it.
This couldn't be farther from the truth, which is that Higgsfield made a promise to open source its product and never did.
For the past two weeks, builders have been recreating popular closed apps as open workflows anyone can run.
So we asked our community to deliver where Higgsfield failed: apps in the open, built by the community, for the community. We've made it clear that the challenge is organized by ComfyUI and not affiliated with or endorsed by Higgsfield.
When we started Comfy, we made the community a promise. Our product stays open, free, and yours. We’ve kept it every day since, and one letter isn’t going to change that.
The challenge stays up. Comfy and open source will win!
2
556
A lego alien just sold me greens gummies...
this is called lego drama style and it might be my favorite ai ad format of 2026
here’s why it slaps:
- Everyone grew up with lego, so it’s instant nostalgia
- It looks nothing like a normal ad, so people don’t skip it
- Brick characters can be dramatic and over the top without being cringe
- You can build ANY world: kitchens, spaceships, alien planets, zero budget
This one’s a concept ad for grüns: girl drowning in veggies → spaceship lands in the backyard → space explorers hand her the secret 🛸
2 years ago this would’ve needed a stop-motion studio and months of work
Now? one afternoon
wanna try the exact tool i made this with?
comment “TOOL” and i’ll send it to you
(spec ad, not affiliated with or endorsed by grüns)
Why is nobody talking about lab-style ads
honestly one of the smartest AI ad formats we've tried this year
what makes it work:
- It hooks people with a question, not a product
- Everyone wants to see who wins the “test”
- The proof comes first, so the product reveal actually lands
- White coats + clean lab = instant credibility
in this spec ad for grüns, scientists put the powder routine through the prep test, then the 30-day test... and the gummies take it home 🧪
normally you’d need a lab set, a cast and a production budget for this
i didn’t need any of it
curious where i made it?
drop “TOOL” in the comments and i’ll DM you the link
(spec ad, not affiliated with or endorsed by grüns)
Everyone's blaming slash today, i don’t think that’s fair
thousands of meta ad accounts went down. we lost 2
same card provider. same BINs. same day
we’ve been with slash since their first 100 customers. still are and still will be.
slash didn’t get your ad account banned. depending on one BIN did.
here’s what actually happens:
virtual cards share BIN ranges. when meta’s risk system gets nervous about a range, it doesn’t check whether you specifically did anything wrong. it shuts off everyone on it. that’s not slash being bad. that’s how payment risk works at meta’s scale, and it can happen to any card provider
and meta isn’t being evil either. scammers burn through card after card, account after account. circumvention policies exist because of them. good advertisers just get caught in the same net
so why did we survive?
we treated payments like infrastructure, not an afterthought
→ spread across multiple BINs, never all on one
→ credit line or corporate card as backup on every account
→ every ad account was advertiser verified
most brands pick their card for the points. 2-2.3% cashback sounds great until your accounts go down mid-Q4 and you lose a weekend of sales
your ad billing shouldn’t be optimised for rewards. it should be optimised for never getting switched off
the points aren’t worth it if the account doesn’t survive
if you got hit today: switch payment method, pay any balance, request a review. do NOT open new accounts to get around it. that’s circumvention and it turns a payment issue into a permanent ban
your card didn’t kill your ad account
your setup did.
Laneige didn't make this ad.
I did. From one product link.
FOOH. UGC. Anime. Claymation. K-drama.
No creators. No shoot. No editing team.
That's 5 of 50+ styles.
Comment "LINK" and I'll run your brand through it.
(Spec ad. Not affiliated with Laneige.)
Dropping $100k on whoever brings me the AI idea that actually breaks ecommerce.
Not an app. Not a chatbot. A paradigm shift.
DMs open.
@manofconnects sent an idea that was actually too good to ignore. Now we’re already beta testing it with brands.
Yeah… that moved stupid fast. People still think execution takes months. With AI the right idea can turn into a working product before most people finish debating it in group chats.
That window is getting smaller every month.
2
231
Swarith retweeted
Even you can generate these kinda videos but the only difference is I didn’t use any prompt:)
Keep figuring it out in the comments 👀
Swarith retweeted
There are 100+ AI UGC tools.
Exactly 0 of them write the prompt for you.
We've been quietly building one.
It researches your product, applies neuroscience to buyer psychology, and writes the prompt that actually converts.
Opening 50 beta seats.
Follow to soon get a chance to get into private Beta tester group
The ad below: made in 2 minutes.
No manual input. No prompt guessing. 100% AI.
🤖 Made with AI
















