Faboo

Data over gut.
Ecom • @voc_labs

[Faboo @faboocsv]

My scroll-depth report said 60% of visitors to one product page stopped under 6%.

I read that as "they scrolled a little."

They never scrolled at all.

On a phone, the page was about 14,300 px tall.
One phone screen is 844 px.

844 ÷ 14,300 = 5.9%.

So "under 6%" = the first screen. Thumb never moved.

Why 5% and not 0%? Clarity seems to count the screen you're looking at.

Over a month on the same page, almost 7x more visits sat at 4–5% than at 0%.

If it only counted scrolling, the people who didn't scroll would pile up at 0%. They piled up at one screen's worth.

Then I lined the pixels up against the page.

What 6 in 10 visitors saw:
→ the photos
→ the title and the stars
→ the top half of one card

What sat below that line:
→ the size picker
→ every price
→ the Add to Cart button

Most of what I'd built on that page sat below the line.

A scroll % is useless until you know how tall the page is.

Convert it to pixels.
Then look hard at the first screen.

For most of your visitors, that's the whole page.

[Faboo @faboocsv]

@arnavsawantt u got it🙌🏻

[Faboo @faboocsv]

Everyone talks about Rosabella's 5,000-word story ads.

So I counted all 864 of their Meta ads.

40% of them are 253 words long.

Same words. Every single time.

One script, 350 ads:

→ 2 versions of the copy. The only difference: one has a tracking link at the bottom
→ ~180 different images on top
→ 168 of those images ran on BOTH versions
→ 276 ads launched on one day. All this script.

It reads like this:

"If your energy is gone by noon, take beetroot."
"If your stamina isn't what it used to be, take beetroot."
"Throw away the detox teas, energy drinks, and trendy supplements. Listen to me carefully."

The other 60%:

→ 98 long stories, ~4,900 words each
→ about 4 ads per story

So it's 2 machines:

→ The stories test the WORDS. 98 scripts, a handful of ads each.
→ The short one tests the PICTURE. Words locked, ~180 images.

And all 864 ads share ONE headline:
"Support stamina and heart health every day."

Most brands change the copy, the image and the headline at once. Then they can't tell what worked.

In Rosabella's short lane, only the picture moves.

[Theo - t3.gg @theo]

Kinda want to get drunk and vibe code until 5am tonight

[Faboo @faboocsv]

@EcomKostnchko_ VAT😭

[Faboo @faboocsv]

@rowaifa affiliate

[Rowaifa @rowaifa]

i'm not supposed to share this but f*ck it

the cheapest way to find seedance is in affiliate discord groups

i'm getting seedance as low as .6 cents per second

kie charges 14.4 cents per second

higgsfield charges 19.6 cents

unless you work for these brands they'll not give you access to the provider

but i'm feeling generous today so

RT + comment "affiliate" and I'll DM you the site

[Faboo @faboocsv]

100% agree! Used to be after. It was in git the whole time and ran for quite a while before I noticed. Definitely had to update some guardrails. Now nothing goes live until I've read the diff.

And don't trust Fiverr reviews on speed gigs. The score jumps, the buyer leaves 5 stars, and most never find out it's fake haha

[Faboo @faboocsv]

I paid a cheap guy on Fiverr to make my Shopify store faster.

He didn't make it faster.

He made it LIE.

15 lines of code, hidden inside a comment that closed early and reopened. So it looked switched off. It wasn't.

What it did:

→ Dropped an invisible box the size of the screen on every page. Speed tools see it and go "wow, loaded already!"
→ Loaded a script from SOMEONE ELSE'S store. It sat quiet until it smelled a speed test. Then it ripped every script off the page.

So Google's test graded an empty page.
My customers got the real one.

I didn't buy speed. I bought a fake report card.

Plus a stranger's code running on a store that takes payments.

I only caught it because I pointed Claude Code at my theme. It found it, traced it, deleted it.

First honest test after:

→ Product page: 53/100
→ 16.6 SECONDS to load the main image on mobile
→ Homepage: 61/100, 12.7 seconds

Brutal. But real.

So:

Don't cheap out on someone touching your theme code.
Or do it yourself. An AI agent will read every line. The $5 guy won't.

2-minute check, right now:

→ View source. Search "GTmetrix", "Lighthouse", "navigator.platform"
→ Search "https://t.co/KYTwp3zel7". Any store number that isn't yours is someone else's code.

If your score jumped after an "optimisation", go look.

[Faboo @faboocsv]

@EcomTable 😭😭😭

[Faboo @faboocsv]

I compared 110 dog supplement ads with 180 Reddit posts from owners of itchy dogs.

Same problem. Two different languages.

The ads explain what's happening inside the dog:

→ "immune" in 64 of 110 ads
→ "gut" in 40
→ "inflammation" in 38

The owners, across all 180 posts:

→ "immune": 3
→ "gut": 1
→ "inflammation": 4

Ads are longer, so I checked per word too. The ads still say "gut" 74x more often.

What owners write instead:

→ "scratching his belly non stop raw"
→ "His back was covered in scabs."
→ "Morning until night she will try to scratch herself whenever possible."

They describe what they can see.

Now the hook. 50 of the 110 ads say "allergy" in their first 100 words.
Owners say it in 26 of 180 posts.

The ads open with the diagnosis.
The owner is still describing the scratching.

The explanation can go in the middle.
The first line should sound like the owner's own post.

Which category should I tear down next? 👇

[Faboo @faboocsv]

I labelled 399 dog skin & coat supplement ads from 9 DTC brands.

110 unique ads, 114,301 words of copy, 8,130 labels.

94% of them talk to the exact same customer.

Here's who, and the one customer almost nobody's talking to.

The ad these brands run, over and over:

→ a regular owner tells the story
→ a long one, about a 4-minute read
→ years of itching that finally ends
→ and someone to blame: vets, big pharma or the supplement industry

The customer they all chase:

→ 3 in 4 ads mention the vet
→ 7 in 10 mention Apoquel
→ they've tried the usual fixes and they're fed up

The burned-out owner on their fifth vet visit.

Same customer. Same story. Same villain.

Now the gap.

Not one ad opens with the coat.

Not shedding. Not shine. Not the smell after a walk.

So I labelled 540 Reddit posts from dog owners.

Most of the shedding posts aren't about the dog. They're about the house:

→ "my couch is not sitable because of all the fur daily."
→ "I can’t wear anything black anymore"
→ "I’m tired of lint rolling the bed every single night!"

In the ads, hair in the home comes up once.

The category is literally called "skin & coat."
Almost every ad sells the skin.

And the owner with hair on the couch is already shopping.
For vacuums, brushes and lint rollers. Not supplements.

Every dog sheds. Not every dog is on Apoquel.

Which category should I tear down next? 👇

[Faboo @faboocsv]

stop using grok bot and start using claude code with proper context management.

[Faboo @faboocsv]

@argooco @karmaco1a honestly just connect meta mcp to claude code.

[Faboo @faboocsv]

who tf is Dr. Sarah Mitchell??😭

[EcomKostnchnko, PHD @EcomKostnchko_]

I wish I could stress how good using Claude Code is for rebuilding your entire theme from scratch, and it's so fucking amazing! It's such a good feeling. You guys definitely need to try it. You just need to learn how to use a text tag properly. It's so fucking beautiful.

[Faboo @faboocsv]

@alexgoughcooper please !

[Faboo @faboocsv]

@jchdotme crazy value🙌🏻

[Mickey @froznato]

This is how we survived recent ban wave 😉

[Faboo @faboocsv]

@vergeecom @jforjacob crazy! did the same for @CarlWeische articles🗣️