Hook variation generator
Genyad can hold the body and call to action of an ad constant and vary only the opening, which is the cleanest creative test available in paid social. The hook decides whether the rest of the ad is watched at all, so it is the variable worth isolating before anything else.
Last updated
How to do it in Genyad
- Pick an ad that already converts the traffic it gets but does not get enough of it.
- Brief 5 to 8 variations that keep the body and CTA and change only the opening clip and first line.
- Ship them as one set and read hook rate, meaning 3 second views over impressions, not conversions.
- Keep the two best hooks and pair them with your two best bodies for the next round.
- Repeat monthly. Hooks fatigue faster than bodies.
What good looks like
A hook test is only readable if nothing else moved. Changing the hook and the offer at once tells you nothing about either.
- Five to eight hooks is enough to find a winner; twenty splits your data.
- Read hook rate first, then completion, then conversion. In that order.
- A hook that wins on hook rate and loses on conversion is attracting the wrong people, which is still useful information.
- Vary the kind of opening, not just the words: a face, a product in motion, a number on screen, a question.
Mistakes that cost you the test
- Judging hooks on conversions. The hook's job is the next two seconds. Judge it on that.
- Changing the hook and the thumbnail frame and the caption style together. Then you have learned nothing.
- Stopping at one winner. Keep two, because the winner will fatigue and you want the runner-up ready.
Frequently asked questions
How many hook variations should I test?
Five to eight in one round. Fewer does not cover enough angles, more splits the budget so thin that no result reaches significance.
What metric shows whether a hook worked?
Hook rate, usually 3 second video views divided by impressions. Completion rate and conversion answer different questions about later parts of the ad.
Should the hook be a person or the product?
Test both. In most accounts a face wins on hook rate and product motion wins on conversion rate, which is exactly why both belong in the set.
How often do hooks fatigue?
Faster than bodies, typically within weeks on TikTok and a month or two on Meta. There is decay data by placement in our 2026 fatigue benchmark.