← All posts
FreeCreativeFor all businesses

Hook rate: the 3-second number that comes before the phone rings

Hook rate shows whether people stop scrolling for your video. How to work it out, what the rules of thumb say, and how to test better openings.

October 29, 20264 min readMadnice Studio

Leads take time. A new video ad can run for a week or more before you know whether it brings in calls. But one number shows up much sooner, and it tells you whether people stopped scrolling at all.

That number is hook rate. Meta doesn't report it under that name. You work it out yourself from two numbers Meta does report.

How to calculate it

Hook rate = 3-second video plays ÷ impressions.

Meta's help center defines 3-second video plays as the number of times your video played for at least three seconds, or for nearly its full length if the video is shorter than that. Impressions is the number of times your ad was shown.

Say your plumbing ad was shown 4,000 times, and it played past three seconds 1,000 times. Your hook rate is 1,000 ÷ 4,000, or 25%. In Ads Manager, add both columns to your report and divide.

What counts as a good number

Meta does not publish an official hook rate benchmark. What you'll find online are rules of thumb from advertisers. Many treat somewhere in the 20s, up to about 30%, as the bar. Many also say to wait until an ad has around 2,000 impressions before you judge it.

Use those figures as starting points, not targets. Your best comparison is your own ads. If your last three videos landed at 18%, 21%, and 24%, a new one at 15% tells you something, whatever the rules of thumb say.

Also remember what hook rate doesn't measure. It says nothing about calls, bookings, or cost per lead. A strong hook with no leads points to the middle of the ad, the offer, or what happens after the click. A weak hook means most people never stayed long enough to hear the offer.

Test hooks on a body that already works

The cleanest test changes only the opening. Here's a simple way to run one:

  1. Pick your best video, the one that has brought in real leads.
  2. Keep everything after the first three seconds exactly the same.
  3. Make three or four new openings.
  4. Run them side by side in the same ad set, with the same offer and the same landing page.
  5. Let each version reach a couple thousand impressions before you compare.
  6. Keep the winner and retire the rest. Try new openings next month.

Because the body stays the same, a big gap in hook rate most likely comes from the opening. One catch: advertisers report that Meta often sends most of the budget to one ad early. If a version barely gets shown, give it more time or test it on its own.

What to change in a hook

Change one thing at a time so you know what made the difference.

  • The opening line. A question, a problem, or a plain fact. Say you run an HVAC company. One version opens with a question like "Is your AC blowing warm air?" Another opens by naming how fast your crew can get there.
  • The first shot. What's on screen in the first second. It could be a technician at the door, the problem up close, like a water stain on a ceiling, or the result, like a family relaxing in a cool living room.
  • The on-screen text. The words that appear in the first seconds, for anyone watching with the sound off. Keep them short enough to read at a glance.

One caution for dental and health practices. Avoid openings that assert or imply the viewer has a condition, such as "Do you have gum disease?" Meta's Personal Attributes policy restricts ads that do that. Lead with the problem or the service, not the person. "Gum disease treatment available" is the kind of framing the policy allows.

New hooks are not new ideas

Hook tests tune an ad. They don't replace fresh concepts.

Meta has said its ad system picks the right creative for each person as businesses upload more varied creative (Meta). In August, Search Engine Land reported a new "Creative diversity" column in Ads Manager reporting. It is marked as in development, and it rates how visually varied your images and videos are as Low, Medium, or High. It's a reporting column, so treat it as a hint, not a grade that punishes you.

Five versions of the same video with different openings will probably look alike to that column. A different reason to buy, a different setting, a different presenter, or a different format is what gives the system something new to work with.

So do both. Each month, add at least one genuinely new concept, and test a few hooks on whichever video is already working. That pairing is how Madnice builds its ads: every concept comes with several hooks, so you can run them side by side and keep the winner.

Sources

Want ads like these, made for your business?

See plans and pricing