PricingLog inOpen eonikGet eonikDownload

Glossary/H

Hook rate

What it meansHook rate is 3-second video plays divided by impressions: the share of a video ad’s impressions in which it played for at least three seconds, read as how often the opening stopped the scroll. Meta does not report it; you add it as a custom metric. One published average, across 88,329 Meta video ads in early 2026, was 25.44%.3

Also called
Thumb-stop ratio, thumb-stop rate
Applies to
Video ads
Set by
The team reading the results, from Meta’s columns
Abinash SenapatiFounder, eonikUpdated Nov 5, 2026Definitions checked against Meta on Oct 26, 2026

The formula

PartMeta’s columnWhat it counts
Top3-second video plays1Plays of at least 3 seconds, or nearly the whole video if shorter; replays not counted
BottomImpressions1Times the ad was shown

Hook rate = 3-second video plays ÷ impressions. The definitions are Meta’s.1

Example numbers. 50,000 impressions and 12,000 3-second video plays give a hook rate of 12,000 ÷ 50,000 = 24%.

Add it in Ads Manager

  1. In Ads Manager or ads reporting, open Columns and choose the Metrics tab.4
  2. Press + Create to open the custom metric form.4
  3. Name it Hook rate and enter the formula: 3-second video plays ÷ Impressions. Show it as a percentage.
  4. Choose whether it is for this report only or reused across your reports. Once a reused metric is shared, its formula cannot be edited, so check it first.4
  5. Add it as a column beside impressions and 3-second video plays, so you can see what it is made of.

Meta works the metric out from the columns in your formula and changes nothing else about them.4 Add ThruPlays ÷ 3-second video plays the same way, as hold rate.

Published figures

IndustryAverage hook rate
All 14 industries25.44%
Toys & games28.79%
Business & industrial28.74%
Health & beauty28.11%
Sporting goods26.89%
Food & beverages26.14%
Home & garden25.06%
Services23.86%
Arts & entertainment23.02%
Apparel & accessories22.94%
Software22.90%
Baby & toddler22.56%
Electronics22.21%
Animals & pet supplies21.32%

Billo’s averages for 88,329 sales-objective Meta video ads with more than 1,000 impressions each, run from January to June 2026, published August 6, 2026. One company’s data, and means, not medians.3

Use a published figure to check that your formula is the common one, not as a target. A different product, audience, placement and length give a different number, so your own past ads in the same placement remain the comparison that means something.

What it means in practice

Compare it with your own ads first, in the same placement and of similar length, over similar numbers of impressions. Because it needs impressions rather than sales, it is also the reading a small testing budget can afford. Wait for enough impressions before reading it; Meta itself does not show its ad relevance diagnostics below 500 impressions.2 A low hook rate, against your own recent ads, points at the opening: the first frame, the first words, the first movement. A high one says only that the opening works.

It says how often people stopped, not whether they liked what they saw, remembered it or went on to act. Read it with hold rate and with the results you care about, never alone.

Write down the formula your team uses, share it, and keep it. Some teams count 2-second continuous video plays instead of 3-second plays,1 and the two give different numbers for the same ad; comparing one with the other compares nothing.

From the number to the next ad

A low hook rate, against your own recent ads in the same placement, says the opening is what to change: the first frame, the first words, the first movement. Keep the body that holds people, and make new openings for it.

Before you write one, look at the openings brands in your category kept running longest in the ad library. A long run means a brand kept paying for that ad; it is never proof the opening works, but it shows which openings survived a market’s own spending.

Then make your version: the same body with new openings, from your own footage and product. How to remake an ad that works covers keeping a structure and making every part your own; in eonik, every ad saved to your moodboard carries Recreate this ad, which keeps its shots, cuts and timing and swaps in your product, your people or your language, with nothing generated until you say so.

Common mistakes

  • Treating a published average as a target.It describes other brands’ ads, not yours.
  • Reading it after a few hundred impressions.It moves a lot with each new day.
  • Treating a high hook rate as success.It measures the opening only.

Sources

  1. About video ad metrics on Facebook and Instagram, Meta Business Help Center. 3-second video plays, ThruPlays and plays at 25% to 100%.Checked Oct 26, 2026
  2. About ad relevance diagnostics, Meta Business Help Center. Quality, engagement rate and conversion rate rankings, and none below 500 impressions.Checked Oct 26, 2026
  3. Video ad hook rate benchmarks, H1 2026, Billo. 88,329 sales-objective Meta video ads with over 1,000 impressions, January to June 2026: an average of 25.44%, by industry; published August 6, 2026.Checked Oct 27, 2026
  4. Create custom metrics in ads reporting, Meta Business Help Center. Columns, the Metrics tab, + Create; a shared metric’s formula cannot be edited.Checked Oct 27, 2026

Related

How to read hook rate and hold rateLearnHold rateGlossaryHookGlossary
On this pageThe formulaAdd it in Ads ManagerPublished figuresWhat it means in practiceFrom the number to the next adCommon mistakesSources

Make your next ad in eonik.

eonik is already on this Mac. Download again

eonik runs on your Mac. Free to start. No card.

Download for Mac

Free to start. No card.Made for Mac. A Windows app is being built.eonik is made for Mac. Free to start. No card.

Product

Download for MacPricingMade in eonikWhat’s new

Use eonik

How to use eonikConnect your assistantCompanion appsAd library

Read

LearnBlogGlossaryCompare

Company

AboutWho it’s forLog inconnect@eonik.ai
© 2026 eonikPrivacyTermsThe Ad Gene communityLinkedInXInstagramTikTok