Hook Rate: How to Calculate It and What Is a Good One
Hook rate is 3-second video plays divided by impressions. A good hook rate beats the median of your own comparable video ads, and you should judge it beside cost per purchase.
By Sachit Sharma, CEO & Founder · Updated 9 Oct 2026
Key takeaways
- 01Hook rate is 3-second video plays divided by impressions. A good one beats the median of your own comparable video ads, because public samples disagree: a mean of 25.44 percent across 88,329 ads in one, a median of 20.9 percent across 479 ads in another.
- 02Hook rate and ad spend are only loosely linked. In one agency's 17-account sample, the five highest-spending video ads in each account had a median hook rate of 22.0 percent, against 20.7 percent for the other ads.
- 03Hold rate is the share of viewers who stay beyond the opening, usually counted with ThruPlays (plays of 15 seconds or more, or to completion). It has more than one formula, so write the denominator next to the name.
- 04Under about 1,250 impressions per ad, do not call a five-point hook rate gap near 25 percent. The worksheet below turns this into a decision.
In this article
- 1How do you calculate hook rate in Meta Ads Manager?
- 2What is a good hook rate on Meta video ads?
- 3Does a high hook rate mean an ad will sell?
- 4Why do two dashboards show different hold rates for the same ad?
- 5How many impressions does a video ad need before its hook rate means anything?
- 6What should you do with a hook rate once you have it?
- 7How do we turn hook signals into a weekly test plan?
- 8Frequently asked questions
Across 88,329 sales-objective Meta video ads that ran in the first half of 2026, the average hook rate was 25.44 percent, according to Billo's benchmark analysis. Billo is a creator-marketing platform, and the figure is one company's sample, not a Meta number.
That caveat matters because hook rate is a ratio you build yourself. Ads Manager, Meta's ad reporting tool, reports the count of 3-second plays and the impressions, and leaves the division to you. Each benchmark therefore reflects who built the formula and which ads they measured.
This page gives you the formula, two public samples that disagree and what that means for your own target, and the impressions you need before a gap is real. It ends with a five-step worksheet that produces one test decision. It is written for anyone looking at a video ad in Ads Manager and deciding whether to keep it, kill it or rewrite its opening.
How do you calculate hook rate in Meta Ads Manager?
Hook rate is 3-second video plays divided by impressions, multiplied by 100. Meta counts a 3-second video play when a video plays for at least three seconds, or for nearly its full length (97 percent) if it is shorter. It excludes time spent replaying the video within a single impression (Meta Business Help: 3-second video plays).
Meta reports the count, so you build hook rate as a custom column in Ads Manager. Many teams call the same ratio thumbstop rate. We use "hook rate" throughout.
Worked example: an ad with 12,000 impressions and 3,000 3-second plays has a hook rate of 3,000 ÷ 12,000 = 25 percent.
Two properties affect how you read it:
- The denominator is impressions, not people. If the same viewer sees an ad five times and only watches once, four impressions add to the denominator with no new play. As frequency rises, hook rate can sag even though the creative has not changed.
- It only exists for video. A static image has no video plays, so it has no hook rate.
What is a good hook rate on Meta video ads?
A good hook rate is one well above the middle of your own comparable ads, because public samples give different answers. Two published samples show how much. Neither is a Meta figure, and both come from companies that sell creative services.
| Sample | Ads and period | Typical value |
|---|---|---|
Billo, a creator-marketing platform | 88,329 sales-objective video ads with more than 1,000 impressions, January to June 2026, $122 million spend, 14 industries | Mean 25.44% |
Biddy, a creative agency: one sample of 17 high-spend accounts | 479 US video ads, each with at least $1,000 spend, from accounts spending $100,000 or more over three months, June to August 2026, $10.1 million spend | Median 20.9%; middle 50% of ads 14.2% to 30.4%; top 10% 44.8% and above |
A mean and a median are different statistics. The median is the middle value, where half of ads score higher and half lower. A mean is pulled upward by a few very high scorers, so the two numbers should not be compared directly. Biddy's two groups also differ: consumer brands (10 accounts) had a median of 22.0 percent and apps and services (7 accounts) 20.0 percent.
Two conclusions follow:
- Segments matter less than accounts. The spread between accounts (account medians from about 13 to 41 percent in Biddy's sample) is much wider than the gap between segments (22.0 versus 20.0 percent). Billo's industry averages run from 21.32 percent for Animals and Pet Supplies to 28.79 percent for Toys and Games, so a pet brand and a toy brand should not share one target.
- Placement changes the number. Biddy notes that an ad delivered mostly in Reels and Stories will usually show a different hook rate than one delivered mostly in lower-cost placements. Reels is Meta's short vertical video format.
Our recommendation: set the bar from your own ads. Take your video ads from the last 90 days in one placement, use their median as the bar, call the top quarter strong and call the bottom quarter candidates for a new opening. Compare ads only against others delivered in the same placement.
Does a high hook rate mean an ad will sell?
Not reliably. Hook rate measures whether the opening held attention for three seconds, not whether anyone wanted the product. In Biddy's sample, the five highest-spending video ads in each account had a median hook rate of 22.0 percent against 20.7 percent for the other ads. Ranking each account's ads by spend and by hook rate, the correlation was close to zero (Biddy).
Our inference: spend is a rough proxy for purchase results, and this study did not measure purchases. The finding is a warning against treating hook rate as a verdict, not proof that it never matters.
An opening can also be startling without being relevant. That raises hook rate and does nothing for purchases, so a high hook rate with weak sales usually means the opening promises something the rest of the ad or the offer does not deliver. Hook signals versus conversion data covers how to weigh the two when budget is at stake.
Hook rate also ignores cost. CPM is the cost per 1,000 impressions. Cost per 1,000 3-second plays equals CPM divided by hook rate, which puts price and attention in one number. This is our arithmetic, not a Meta metric:
| CPM | Hook rate | Cost per 1,000 3-second plays |
|---|---|---|
| $30 | 20% | $150 |
| $30 | 30% | $100 |
| $40 | 10% | $400 |
| $200 | 50% | $400 |
A 50 percent hook rate at a $200 CPM buys attention at the same price as a 10 percent hook rate at a $40 CPM. If your hook rate is excellent and your costs are still high, check the audience before the creative. Narrow age or gender targeting and expensive categories both raise CPM.
Why do two dashboards show different hold rates for the same ad?
Because hold rate has no single formula. A ThruPlay is Meta's count of video plays that reached completion or at least 15 seconds (Meta Business Help: ThruPlays). Teams then divide ThruPlays by either impressions or 3-second plays.
Biddy's sample divides by impressions and reports a median hold rate of 4.6 percent, with the middle half of ads between 3.0 and 7.3 percent. It also reports that a median of 22 percent of 3-second viewers reached a ThruPlay. That is the same data under the second formula.
An illustrative ad with 12,000 impressions, 3,000 3-second plays and 660 ThruPlays shows the gap:
- ThruPlays ÷ impressions = 660 ÷ 12,000 = 5.5 percent
- ThruPlays ÷ 3-second plays = 660 ÷ 3,000 = 22.0 percent
Neither is wrong. A threshold only works with the formula it was built on, so write the denominator into the column name. An ad shorter than 15 seconds can reach a ThruPlay by finishing, while a 45-second ad needs 15 unique seconds, so compare hold rate between ads of similar length.
How many impressions does a video ad need before its hook rate means anything?
About 1,250 impressions per ad to separate 25 percent from 30 percent, and far more for smaller gaps. Hook rate is a proportion, so its uncertainty shrinks with volume. This table is our model, using a 95 percent confidence interval for a hook rate near 25 percent: margin = 1.96 × √(0.25 × 0.75 ÷ impressions).
| Impressions on the ad | Margin of error |
|---|---|
| 1,000 | ±2.7 points |
| 2,000 | ±1.9 points |
| 5,000 | ±1.2 points |
| 10,000 | ±0.85 points |
| 20,000 | ±0.6 points |
To compare two openings, we used the standard two-proportion sample size at 5 percent significance and 80 percent power: n per ad = (1.96 + 0.84)² × [p1(1−p1) + p2(1−p2)] ÷ (p1 − p2)².
| Gap you want to detect | Impressions needed per ad |
|---|---|
| 25% vs 35% | about 330 |
| 20% vs 25% | about 1,090 |
| 25% vs 30% | about 1,250 |
| 25% vs 28% | about 3,400 |
The model assumes every impression is independent. Repeat impressions to the same person break that assumption, so treat the numbers as minimums.
The same formula shows why hook rate settles earlier than a purchase metric. Separating a 2.0 percent purchase rate from a 2.5 percent one needs about 13,800 observations per ad, because 7.84 × (0.02 × 0.98 + 0.025 × 0.975) ÷ 0.005² ≈ 13,800. Each observation is a click or visit, which takes many impressions to collect. Hook rate works for early triage and not for final verdicts.
What should you do with a hook rate once you have it?
Run five steps, then write one decision. Each step has a pass or fail you can check.
Build and label the columns
Name them "Hook rate (3-second plays ÷ impressions)" and "Hold rate (ThruPlays ÷ 3-second plays)", or whichever denominator you choose. Pass: every column states its formula. Fail: any unlabeled ratio.Check volume
Look up your gap in the impressions table above. Pass: each ad you are comparing has at least that many impressions. Fail: keep running the ads and make no call.Place each ad against your own account
Take your video ads from the last 90 days in the same placement and find the median and the lower and upper quartiles. Label each ad strong (top quarter), middling or weak (bottom quarter), and do the same for hold rate. Pass: every ad has a label. Fail: fewer than about 20 comparable ads, in which case use the ad's own history and wait. The 90 days and the 20-ad floor are our recommendation.Pair it with cost
Find the ad's row in the table below. Pass: every ad sits in one row. Fail: any ad with no row, in which case recheck its step 3 label.Write the decision as one sentence
"Change ___ in the first three seconds, keep ___ constant, to learn ___." Keep the control the same, as in choosing a creative test control ad. Pass: the sentence names one variable changed and one held constant. Fail: it names two or more changes.
| Hook rate | Hold rate | Cost per purchase vs your target | What to do |
|---|---|---|---|
| Weak | Any | At or under target | Keep running it. A converting ad is not a failure. |
| Weak | Any | Over target | Rewrite the opening and keep the body, offer and landing page the same. |
| Middling | Any | At or under target | Keep running it. |
| Middling | Any | Over target | Check hold rate and the offer first. Rewrite the opening only if hold rate is also weak. |
| Strong | Weak | At or under target | Keep running it, and note that the opening may be losing viewers after the hook. |
| Strong | Weak | Over target | The opening promises more than the rest delivers. Align the first three seconds with the product. |
| Strong | Middling or strong | Over target | The problem is past the creative opening: offer, price, landing page or audience. |
| Strong | Middling or strong | At or under target | Test variations of this opening and scale within your budget rules. |
If CPM is unusually high, compute CPM ÷ hook rate before judging the creative.
No single metric should kill an ad. A workable rule is to pause an ad that spends well past your target cost per purchase without a conversion, and to use hook rate to decide what to rewrite.
How do we turn hook signals into a weekly test plan?
We treat hook rate as a pointer and a ranked test as the decision. Deepsolv's creative intelligence platform gives Meta advertising teams a ranked weekly creative test plan that shows what to test, what to skip and why. It draws on ad-performance data, competitor activity, customer signals and past ad learnings, and each test comes with an execution-ready creative brief.
Our competitor tracking covers hooks, messaging, offers and formats, so you can see which openings competitors keep running before you choose your next variable. We also keep a record of what worked and what failed, so the next test starts from your history. The missing layer between a metric and a decision is covered in why creative analytics is not a test plan.
To check an opening before you spend on it, export its first frame as a still image and upload it to our free Attention Heatmap, which predicts where viewers' eyes will land. It takes JPEG, PNG or WebP images and needs no signup. It shows predicted attention on a frame, not your hook rate, and how to read an attention heatmap explains when to trust the result.
To see a ranked plan built on your own account, book a personalized demo with our founding team, and ask about pricing there.
