Pulsar · Field guide · No. 22 · Why the scroll is the real competitor, the hook types that stop it, and how to measure and iterate the opener.
The First 3 Seconds
You don’t lose people at the end of the ad — you lose them in the first three seconds. The hook is most of your performance. Here’s how to build one.
Your competitor is the thumb
You’re not really competing with other advertisers — you’re competing with the scroll. A feed is an endless, faster-than-thought flick, and your ad has a heartbeat’s worth of time to earn a stop.
Everything after the hook only matters to the people the hook kept. Lose them at second two and the best offer in the world never gets read.
What a hook actually does
Two jobs, both inside three seconds: stop the scroll (a pattern break) and promise a reason to stay (a payoff). Miss either and the thumb keeps moving.
Hooks that stop the scroll
A handful of openers do most of the work. Reach for one deliberately.
| Hook type | Example opener |
|---|---|
| Pattern break | an unexpected visual or motion in frame one |
| Bold claim | “Your dashboard is lying to you about sales.” |
| Relatable problem | “Ever scaled a winning ad and watched it die?” |
| Curiosity gap | “One column in Ads Manager quietly costs you money.” |
| Show the payoff first | the result, up front, then how |
Say it, don’t bury it
Put the payoff up front. No five-second logo sting, no slow intro, no throat-clearing. The most common hook mistake is spending the three seconds you have earning the right to start.
Match the hook to the audience temperature
Cold audiences need a stronger pattern break — they owe you nothing. Warm audiences (retargeting) already know you, so you can be direct: the reminder, the answer, the offer.
Measure the hook, not the whole ad
The hook has its own metric: 3-second video plays / thumbstop rate / hook rate (how many who saw it kept watching). A low hook rate caps everything downstream — no CTR or CVR can recover an ad nobody watched.