How to improve hook rate on Meta video ads

Hook rate is the share of impressions that stay for three seconds. Raise it by changing the first frame, the first line and the first promise, one at a time.

What hook rate is and how to calculate it

Hook rate is 3-second video plays divided by impressions. Ads Manager gives you both numbers. Multiply the result by 100 and you have a percentage. Meta does not show it as a default column, so most teams add it as a custom metric.

Example, not a benchmark: an ad gets 20,000 impressions and 5,000 three-second plays. Hook rate is 25%. A different ad gets 20,000 impressions and 7,000 three-second plays. Hook rate is 35%. The second ad is better at stopping the scroll. That is all the number says.

Hold rate is the next step. Divide ThruPlays (or 15-second plays, pick one and stick with it) by 3-second plays. Hook rate tells you whether people stopped. Hold rate tells you whether the body kept them. Keep the two apart, because they have different fixes.

A Meta hook rate benchmark you can trust

There is no honest universal number. Hook rate shifts with placement, video length, audience temperature, how autoplay starts and how Meta counts a play. A figure from someone else's account will mislead you more than it helps.

Build your own benchmark instead. It takes twenty minutes.

The method is simple. Pull your last 20 to 30 video ads that each got a meaningful amount of delivery. Sort by hook rate. Note the median and the top quarter. The median is your baseline. The top quarter is your target. Do this per format (UGC talking head, product demo, motion graphic) and per audience type (prospecting vs. retargeting), because they will not sit in the same range.

Then read the benchmark in context. A hook rate below your median on a cold audience is a hook problem. A high hook rate with no sales is not a win. It may be a hook that attracts the wrong people, which we cover below.

Diagnose before you rewrite the hook

Low hook rate is only one of four places an ad can leak. Fix the one that is actually broken.

Read the funnel left to right and stop at the first weak step.

• Hook rate low: the first 3 seconds fail. Change the opening frame, line or promise.
• Hook rate fine, hold rate low: the opening promise is not paid off, or the pacing sags after the hook. Rework seconds 3 to 15.
• Hold rate fine, click-through low: the offer or call to action is weak, or the promise is not clear by the end.
• Click-through fine, conversion low: look at the landing page and whether it continues the ad. This is not a hook problem.

Seven things to change in the first three seconds

Most hook work comes down to a short list. Change one item per variant so you can see which one moved the number.

• The first frame. Open on motion, a face mid-expression or a surprising object, not a logo card or a slow fade. Many people see a still before the video plays.
• The first spoken line. Start mid-thought. Skip greetings and introductions. Say the problem, the result or the odd claim in the first sentence.
• On-screen text. Put the hook in plain words on screen in the first second. Many viewers watch without sound. Keep it short enough to read in two seconds.
• The promise. Name a specific outcome, person or situation. A hook for 'people who wake up at 3 a.m.' beats a hook for 'anyone who wants better sleep', because the right person recognizes themselves.
• The awareness stage. A cold audience that does not know the problem needs a different opener than people who already compare products. Match the hook to what the viewer already knows.
• The visual proof. Show the thing happening: the stain disappearing, the product in a hand, the before state. A visible result in the first seconds beats a description of it.
• The crop. Check the hook in 9:16 and 4:5. Captions and faces that sit too low or too close to the edge get covered by interface elements in Stories and Reels.

How to test hooks so the result means something

Keep the body of the video identical and swap only the first three to five seconds. That isolates the hook. If you change the hook, the offer and the creator at once, you learn nothing.

Run the variants in the same ad set so Meta delivers to the same audience under the same conditions. Three to five hooks per concept is enough. More just splits delivery into thin slices.

Do not call a winner early. A hook that looks strong after a few hundred impressions often settles back to the pack. As a rule of thumb, wait until each variant has several thousand impressions before you compare hook rates, and treat small gaps as noise.

Name every variant so you can read the results later. A simple pattern works: batch number, concept, hook number, for example B14-ConceptA-H3. Without it, a month from now nobody remembers which opener was which.

Keep a log of what each winning hook did. Write one line: the desire it spoke to, the angle, the awareness stage, the format. After ten or fifteen tests, patterns show up that no single ad would reveal.

Mistakes that raise hook rate and lower sales

Clickbait hooks. A shocking opener that has nothing to do with the product earns 3-second plays from people who will never buy. If hook rate rises while ROAS or CPA gets worse, the hook is attracting the wrong audience.

Five hooks with the same promise. Different words, same claim. You tested one idea five times. Vary the promise, not just the phrasing.

Judging by hook rate alone. It is a diagnostic, not a verdict. An ad with an average hook rate can still carry a campaign if it converts. The verdict is spend and return.

Comparing across formats. A static-style motion ad and a UGC talking head will never share a baseline.

Copying a hook that worked once. A hook wears out as frequency climbs. Plan the next set before the current one fades.

Skipping the landing page. If the page does not continue the promise in the hook, you pay for attention and lose it on the click.

From hook rate to a verdict on the batch

Hook rate tells you why an ad did or did not stop people. It does not tell you whether the concept made money. For that, you need a rule written before the results come in.

Hitrate plans each test as a numbered batch, one concept in one ad set, and writes the ad names for you so every hook variant maps back to its batch, desire, angle, awareness stage, format and creator. After 7 days it labels the batch Scaler, Spender, Efficient or Miss by one written rule, per campaign. Then it shows hit rate by desire, angle, awareness and format, so you see which kinds of hooks keep turning into spend.

You can run the same logic in a spreadsheet. The point is the same: a hook is a hypothesis, the batch gets a verdict, and the next brief starts from what sold.

Related guides

• The creative strategist workflow, from research to learnings
• How to identify creative fatigue in Meta ads (with thresholds)
• Meta hook rate benchmark: what counts as a good hook rate
• A Meta ads creative testing framework you can run weekly
• When to Kill a Meta Ad in a Creative Test

Updated 2026-10-02