Hook rate tells you which part of an ad is broken. It doesn't tell you whether the ad is working.
That distinction matters. Most accounts we audit either ignore hook rate completely, or treat it like the scoreboard and kill ads that were quietly making money.
This guide covers the formula, how to read hook rate alongside hold rate, the hook formulas we use, and how to test hooks so you know which one actually won.
What is hook rate?
Hook rate is the share of people who saw your video ad and kept watching past the first 3 seconds. It measures one thing: did the opening stop the scroll?
Hold rate is its partner. It's the share of people who kept watching to 15 seconds. It measures whether the rest of the ad paid off the promise the hook made.
You'll also see hook rate called thumbstop ratio. It's the same idea: of everyone who saw the ad, how many stopped.
The hook rate formula
Hook rate = 3-second video plays ÷ impressions
Hold rate = ThruPlays ÷ impressions
For example, if an ad gets 20,000 impressions and 6,000 people watch past 3 seconds, the hook rate is 30%. If 3,000 of them reach 15 seconds, the hold rate is 15%.
How to see hook rate in Meta Ads Manager
- Open Ads Manager and switch to the ad level. Hook rate only makes sense per ad.
- Go to Columns, then Customise columns.
- Add Impressions, 3-second video plays and ThruPlays.
- Create a custom metric for hook rate (3-second video plays divided by impressions) and one for hold rate (ThruPlays divided by impressions).
- Save it as a column preset so it's there every time you review creative.
What is a good hook rate?
It depends on the account, the format and the placement, which is why we don't judge ads against a universal number.
Format alone moves it. In our own testing, changing the format of an ad has shifted hook rate by roughly 10 to 15% while CPA stayed flat. So a hook rate that looks "low" for a static-style video might be perfectly healthy, and a "high" one might be doing nothing for sales.
The useful comparison is inside your own account: the same body, the same format, different hooks. That's where hook rate tells you something real.
What a genuinely better hook looks like: one brand came to us running generic UGC. We shot a 60-second founder origin story with the same product and the same offer. Hook rate went from 28% to 52%.
Hook rate diagnoses. It doesn't decide.
CPA and spend decide whether an ad lives. Hook rate tells you which part of the ad to fix.
Here's how we read it:
| Hook rate | Hold rate | What it means | What to do |
|---|---|---|---|
| Strong | Weak | The hook works, but the body doesn't pay it off | Rewrite the middle |
| Weak | Strong | The right people stay, but too few stop | Rewrite the first 3 seconds |
| Strong | Strong, but CTR is weak | People watch, but don't act | Fix the offer or the CTA |
| Weak | Weak | Nothing is landing | Kill it |
Why we're strict about this: in one account we audited, hook rate had a negative correlation with ROAS (−0.216) across 34 funded videos. That doesn't mean low hook rates cause sales. It means hook rate on its own wasn't predicting revenue in that account, and anyone optimising for it alone would have cut the wrong ads.
The other trap is bait. A hook that grabs attention but isn't relevant gets a high hook rate and a collapsing hold rate. A hook is a promise of relevance, not a trick.
The hook formulas we use
1. The problem hook. Name a specific frustration in 3 seconds. "Hate that chalky, fake-sweet aftertaste most protein bars leave?" Use it when the pain is widely felt. It's strong on warm audiences.
2. The curiosity hook. Open a loop the viewer wants closed. "What if I told you the best-performing ad for this protein bar doesn't even show the product?" Best for cold audiences.
3. The pattern interrupt. Something unexpected on screen. Someone spitting out a competitor's protein bar: "Yeah. We hate that too." Use it for strong visual demos or crowded markets.
4. The contrarian hook. Disqualify people on purpose. "Don't call us if any of this sounds like you."
5. The identity hook. Speak to who the viewer is. "There's always one mum in the group chat who knows."
The specificity rule applies to all of them. "Finally, a white t-shirt that doesn't shrink into a belly top after one wash" beats "Check out our new t-shirt" every time.
How to test hooks properly
The only way hook rate teaches you anything is if the hook is the only thing that changed.
Build a three-hook cell. One problem hook, one curiosity hook and one pattern interrupt, all on the identical body, product, CTA, offer and targeting. Only the first 3 seconds change.
Keep the setup simple. One campaign, the 3 ads in one ad set. Don't move budget until the read window closes, and don't judge on CPA before you have a signal on hook rate.
Shoot more hooks than you need. When you film an ad, film six hooks. It costs almost nothing on the day and gives you two full test cells.
Give each hook enough budget to mean something:
| Monthly account spend | Test budget per ad | Read window |
|---|---|---|
| Under $5k | $15 to $25 a day | 5 days |
| $5k to $30k | $30 to $50 a day | 3 to 5 days |
| $30k to $100k | $75 to $150 a day | 2 to 3 days |
| $100k+ | $200 to $500 a day | 2 days |
Log every test. For every cell, write down the hook framework, the hook word for word, hook rate, hold rate, CTR, CPA, and one line on why it won. After six to eight weeks, you have your own hook database, and you stop guessing.
7 common hook mistakes
- Leading with what you're proud of. "Introducing our new protein powder" answers a question nobody asked. The viewer is asking "is this for me?"
- Being clever instead of clear. "Finally, a dress with actual pockets" beats a pun.
- Running too long. A hook is 3 to 5 seconds. If it needs longer, it's the body.
- Opening with the brand story or the logo. Earn attention first.
- Testing hooks with different bodies underneath. If hook two wins, you can't tell if it was the hook or the edit.
- Writing swappable hooks. If you can swap another brand's name in and it still makes sense, it's a tagline, not a hook.
- Chasing hook rate with bait. Hook rate goes up, hold rate collapses, and CPA doesn't move.
How we use hook rate for clients
Hook testing is built into every ad we make. Every video goes out with several hooks on the same body, and every result goes into a tracker, so our clients always know which hook won and why. It's the core of our creative testing system, and it's why the founder story above could be rebuilt, iterated and scaled rather than found once by luck.
If you'd rather have hooks written, shot and tested for you, that's what our video and static ad and UGC ad teams do every week.
Want to know which of your hooks are actually working?
Book a Creative Diagnostic. It's a live teardown of your account's creative on the call: which hooks are carrying your spend, which are just grabbing attention, and what we'd test next.
Book a free Creative DiagnosticA live teardown of your ad creative, on the call. No obligation.FAQs
What is hook rate?
Hook rate is the percentage of people who saw your video ad and watched past the first 3 seconds. It shows whether the opening of the ad stopped the scroll.
How do you calculate hook rate?
Divide 3-second video plays by impressions. For example, 6,000 three-second plays from 20,000 impressions is a 30% hook rate. You can add it as a custom metric in Meta Ads Manager.
What is a good hook rate?
There's no universal number. Format, placement and audience all move it. Compare hooks inside your own account on the same body and format, and judge the ad on CPA, not hook rate alone.
What's the difference between hook rate and hold rate?
Hook rate measures who stopped in the first 3 seconds. Hold rate (ThruPlays divided by impressions) measures who kept watching to 15 seconds, which shows whether the body of the ad paid off the hook.
Is hook rate the same as thumbstop ratio?
Yes. Both describe the share of people who stopped scrolling and watched the first few seconds of your ad.