Facebook Ads Management

How I validate if an ad is actually a winner before scaling it.

It's really not about ROAS or CPA. It's about purchase volume + time.

My validation threshold:

→ The ad needs at least 3-4 purchases per day for at least 3 days before I make a decision

Why both purchase volume

I&#39;ve audited 150+ Meta Ad accounts this year.

These 5 critical structural issues appeared in almost EVERY account:

1. 𝗡𝗼 𝗰𝗹𝗲𝗮𝗿 𝘀𝗲𝗽𝗮𝗿𝗮𝘁𝗶𝗼𝗻 𝗯𝗲𝘁𝘄𝗲𝗲𝗻 𝘁𝗲𝘀𝘁𝗶𝗻𝗴 𝗮𝗻𝗱… | Toby W. | 13 comments

Not to dunk on Olly but these are actually considered very similiar ads in Meta's system. 

We recently got an audit from our Meta reps and I was surpised to see what was considered different. 

It's not just visuals. It's the whole package; visual, copy, offer, and most

This is my client's testing and scaling structure

CBO - Scaling
CBO - Testing (w/ min spend limits)

Let me explain what's unique about that

The scaling CBO has about 12 different ad sets

And here's why..

We tried consolidating all "proven winners" into a few ad sets (1 ad

A profitable Meta ad account starts with the right foundations.

Because most accounts fail in the system behind it.

Here’s how I break down every account before scaling:

- Consolidation → Are you… | Travis Moh☘️ | 60 comments

A diagram titled <n>What Makes a Profitable Ad Account?</n> by <n>Travis Moh</n>.

How to make optimizations in your account without breaking what's working.

When to turn ads off:

→ After 3 days of consistently bad performance
→ After spending 3-4x your average CPA per day for those 3 days
→ Check add to cart rate - if it's low, people aren't entering the

Here's my testing process

1. Launch a concept per ad sets (max 3 variations)
2. Let it run for a min of 3 days
3. Let an individual ad spend 3 times the breakeven CPA a day.

Basically the ad you're thinking about turning off must have been running for at least days, and have

Never launch 12 ads around 1 angle.

Launch 12 ads across 12 angles.

Each ad =

→ 1 unique desire
→ 1 awareness stage
→ 1 avatar
→ 1 sophistication level

Let Meta test everything.

Once it picks a winner → make 12 more like it.

My account structure forces Meta to spend on new ads instead of letting old winners hog all the budget.

Here's how:

My scaling campaign always has one ad set with all the historical winners in one place.

→ I set a max spend limit on that ad set so those winners don't eat all