Most beginners run Facebook ads like it’s 2019. They spend hours picking interests, building lookalikes, and splitting audiences into tiny groups.
Here’s what changed: Meta’s algorithm is now better at finding buyers than you are. Broad targeting often works best because it gives the algorithm room to learn.
So if targeting isn’t the lever anymore, what is?
Your creative is your targeting. The video itself tells Meta who to show it to.
Here’s how I approach it:
1. The first 3 seconds are everything ⏱️
If the hook doesn’t stop the scroll, nothing else matters. Open with the problem, a bold claim, or the product doing something unexpected. Never open with your logo.
2. Test hooks, not whole new videos 🔁
Take one solid video and swap only the first 3 seconds. Make 3 to 5 versions. This is the fastest, cheapest way to find what works.
3. Set a kill rule before you launch ✂️
Decide in advance how much you’ll spend before cutting an ad. A simple starting point: if an ad spends about 1.5x your product price with zero sales, cut it. No emotions, no “let’s give it one more day.”
4. Scale the winner, not the hope 📈
When one creative clearly outperforms, put your budget behind it and make new variations of that angle.
Bonus sauce 🔥
Give ChatGPT this prompt:
“I’m selling [product] to [customer]. The main problem it solves is [problem]. Write 10 scroll-stopping video hooks under 10 words each. Mix problem-based, curiosity-based, and bold-claim hooks.”
Now you’ve got 10 hooks to test, and it took 30 seconds.
Inside the Classroom, I walk through exactly how we build, test, and scale ad creatives with AI.
👇 Comment below: What’s the best hook you’ve ever seen in an ad? The one that actually made you stop scrolling.