Skip to content

Video

Hooks for video ads: how to stop the scroll

You get about a second. Use it.

Hooks for video ads are the first second or two of the ad: the opening shot, line and sound that decide whether someone keeps watching or scrolls past. The best hooks show the result, name the problem or make a surprising, true claim straight away, and the only way to know which works for you is to test several against the same ad.

Robert S., Head of PR & Content, Foundry Agency

By Robert S.Head of PR & Content · 7 min readUpdated

What are hooks for video ads?

A hook is everything the viewer sees and hears in the first moments of your ad. Its only job is to earn the next few seconds.

It's three things working together: the opening shot, the words on screen or spoken, and the sound. A great line over a dull shot of a logo won't stop anyone. A striking shot with no context won't either. The best hooks get all three pointing the same way.

Why does the hook matter so much in an ad?

Because nobody chose to watch your ad. On a feed they're scrolling past it. On YouTube, skippable ads can be skipped after 5 seconds. If the opening doesn't land, the rest of the ad never gets seen, however good it is.

The hook also decides who stays. A hook that speaks to the right person filters out the wrong ones, which means fewer wasted clicks and a better chance the people who keep watching will buy.

What types of hook work in video ads?

These are the ten we use most. Most ads can be opened in three or four of these ways from the same footage.

  • The result first: the finished job, the transformation or the product working.
  • The problem: show or say the frustration your buyer has, in their words.
  • The call-out: name the audience, such as cyclists in London or bid managers.
  • The surprising fact: something true about your product the viewer wouldn't expect.
  • The question: one the viewer is already asking themselves.
  • The before and after: two states, side by side or cut together.
  • The demo in motion: hands on the product, already moving.
  • The reaction: a real customer's face at the moment it clicks.
  • The contrast: what most people do, then what you do differently.
  • The pattern break: an unexpected shot, sound or angle that doesn't look like an ad.

How do you write a hook for a video ad?

Start from the buyer, not the brand. These steps work for any product.

  1. Write down who the ad is for and the one problem it solves for them.
  2. List the moments in your footage that show that problem or its result most clearly.
  3. Write five to ten opening lines, each under ten words, using different hook types.
  4. Pair each line with the shot that makes it strongest.
  5. Check every hook is true and matches what the rest of the ad delivers.
  6. Cut three or four of them onto the same ad body, so the hook is the only thing that changes.
  7. Run them against each other, and keep the winner as the baseline for the next round.

How should a hook look and sound?

It should make sense in a single frame, with the sound off. Put a short line of text on screen in the first moment, in large type, inside the platform's safe zones so the app's buttons don't cover it.

Move straight away. Start mid-action, make the first cut quickly and avoid fades from black. If there's a voice, start on the strongest word, not on a breath or a greeting. Put your brand in early, but don't open on a logo card.

How do you test hooks properly?

Change one thing at a time. Keep the body of the ad the same and swap only the first few seconds, so you know the hook made the difference. Meta and TikTok both have built-in tools for testing ad versions against each other.

Judge on what matters. A hook that gets people watching but brings the wrong people isn't a winner. Look at clicks, leads and sales as well as how long people watch. Then write new hooks in the style of the winner, because every hook tires eventually.

What hook mistakes should you avoid?

We see these in most ad accounts we take over.

  • Opening on a logo, a greeting or a slow pan.
  • Clickbait that the rest of the ad doesn't deliver. It wastes budget on people who won't buy.
  • Claims you can't back up. UK ads must not mislead, and the ASA can ask for proof.
  • A hook that only works with sound on.
  • Text placed where the app's buttons or caption will hide it.
  • Testing a new hook and a new offer at the same time, so you can't tell which worked.

How does Foundry write and test hooks?

We edit video ads and run the campaigns too, so hooks aren't a guess. Every ad we cut comes with several openings on the same body, captioned and sized for each placement, ready to test.

When the results come in, we keep the winner, write the next round in its style and retire the tired ones. Meta campaigns take about 4 hours to set up on average, so testing starts fast.

Our averages. Not our best days.

Our averages

  • 3hours

    Google Ads setup

  • 4hours

    Meta Ads setupthis service

  • 20mins

    Average reply time

Video ad hook FAQs

  • Think of the hook as the first second or two. It should make sense in a single frame, then the next few seconds have to keep the promise it made.

  • Start with three or four on the same ad body. That's enough to see a difference without splitting your budget too thinly.

  • Yes, for most feed ads. Plenty of people scroll with the sound off, so a short line of text carries the hook for them.

  • There's no single best one. Showing the result first works for many products, but the only way to know for yours is to test it against other types.

  • For a while. Every hook tires as the same people see it. Write new ones in the style of the winner before results drop.

  • No. A hook the ad doesn't deliver on brings clicks that don't buy, and a misleading claim can break UK advertising rules.

Want ads that stop the scroll? Get moving.

Tell us what you need. We'll reply within the hour (working hours), and most websites go live about 5 days after kick-off.

We'll reply within the hour (working hours).

Or email hello@foundryagency.co.uk

Analytics cookies are off until you accept. We use them to see which pages help. Cookie notice