How to test ad hooks: one body, five openings

Guide · 5 min read · By Roshni at Weblur · 11 Nov 2026

The hook is the first few seconds of your ad. It is the part that decides whether anyone stays. And it is the easiest part to test.

This is the method I use when I recut an ad: keep the body, change the opening.

What a hook actually is

A hook is whatever the viewer sees and hears before they decide to keep watching or scroll. It can be a line of text, a spoken question, a striking visual, or the most interesting moment pulled to the front.

It is not the whole ad. It is the doorway.

Why test hooks instead of whole ads

When you change the whole ad, you do not know what made the difference. When you change only the opening and keep everything else, you do. The body stays constant, so any change in results points back to the hook.

It is also cheaper. The footage already exists.

The method: one body, several openings

1. Pick an ad that already works. The body should be solid.
2. Write down what the viewer needs to feel in the first second. Curiosity? Recognition of a problem? A promise?
3. Make a handful of openings, each with a different angle.
4. Keep everything after the hook identical.
5. Run them side by side and compare.

Different angles to try

• Problem first: start with the pain. ("Losing your hair?")
• Result first: start with the outcome. ("This stopped my hair falling out.")
• Curiosity: raise a question the body answers.
• Strongest moment: pull the best moment of the footage to the front.
• Direct address: speak to one specific person.

Each angle speaks to a different kind of viewer, which is why testing beats guessing.

Keep the test fair

• Change only the opening. Same body, same music, same captions after the hook.
• Let each version run long enough to see a pattern, not one lucky day.
• Compare like with like: same placement, same audience.
• Write down what you tested, so you do not repeat it later.

What to look at

Watch how many people stay past the opening, and how that carries through to clicks and results. Your ad dashboard has the numbers. One number alone rarely tells the story, so look at them together.

A small example

A recut I did on a live ad used two openings on the same body: one started with the problem, one started with the result. Same footage underneath. That is all it took to get two genuinely different ads from one.

(Note: this was a concept recut of a public ad, not a paid client project.)

Summary

• The hook decides whether people keep watching.
• Test openings, not whole ads: keep the body, change the first seconds.
• Try different angles: problem, result, curiosity, strongest moment, direct address.
• Keep the test fair and write down what you learn.