Hook Rate and Hold Rate on Meta: Formulas, Benchmarks and Fixes
Hook rate is not one of Meta's own metrics. Its list of video ad metrics has no entry for it, so you build the number from inputs Meta does report. The same goes for its partner, hold rate.
That is where the confusion starts. Guides disagree on the hold rate formula, and the benchmarks quoted online come from agencies and tool makers, not from Meta. A score that looks strong in one account can be ordinary in another.
You will get both formulas from Meta's own definitions, a worked example, and the edit for each weak spot.
What Is Hook Rate on Meta Video Ads?
Meta reports every input you need, though not the ratio. Hook rate is the share of impressions that turned into a 3-second video play. It tells you how often your opening stopped someone mid-scroll.
A 3-second video play counts each time your video plays for at least 3 seconds. Shorter videos count once they play nearly in full. Time spent replaying inside one impression is left out, per Meta's guide to video ad metrics.
An impression counts each time your ad is on screen for the first time. A video does not have to start playing for an impression to count. So a fast scroll past still lands in the denominator.
Some tools call the same number thumbstop rate or thumbstop ratio.
Hook Rate Formula
Hook rate = 3-second video plays ÷ impressions × 100
Say a product demo gets 24,000 impressions, and 7,440 of them turn into 3-second plays. Divide 7,440 by 24,000 and you get 0.31, a 31 percent hook rate.
Meta also reports 2-second continuous plays, but the guides we checked use the 3-second count.
Hold Rate Formula
Hold rate = ThruPlays ÷ 3-second video plays × 100
It shows how many 3-second viewers stayed for a ThruPlay.
A ThruPlay is a video played to completion or for at least 15 seconds. Meta's metric page now calls them 15-second ThruPlays. For videos under 15 seconds only a full watch counts, and Meta treats 97 percent of the length as full.
Guides split on the top of that fraction. Most use ThruPlays or 15-second plays, but Hawky's definition uses completions, which run lower on videos over 15 seconds.
Length skews it too. A ThruPlay on a 12-second video is a full watch, but on a 45-second video it is the first third. Compare hold rates between videos of similar length.
How to Calculate Hook Rate and Hold Rate in Ads Manager
Work at the ad level, because each video is its own test. Open the Ads tab, click Columns, then Customize columns, and tick these metrics:
- Impressions and link clicks
- 3-second video plays and ThruPlays
- Video plays at 25, 50, 75, 95 and 100 percent
Check Save as a column preset, then click Apply, so the set waits under Columns next time. Pick the last 14 to 30 days, then export the table.
Then work out both rates for each ad. For the video average, divide the account's total 3-second plays by its total impressions. Averaging the ad percentages would give a small ad as much weight as your biggest one.
Hold off judging any ad under 2,000 impressions. A few plays either way swing its rate by whole points.
What Is a Good Hook Rate for Meta Ads?
No benchmark comes from Meta, since hook rate is not one of its metrics. Agency and tool blogs fill the gap, and the three below all start near 25 percent.
| Publisher | Date | Hook Rate Ranges | Hold Rate |
|---|---|---|---|
| Segwise | March 2026 | 25 to 30% good, 30%+ for top creatives | Not given |
| Hawky | March 2026 | 25%+ table stakes, 30%+ good, 40%+ elite | 40 to 50% target |
| Exposure | June 2026 | 25 to 35% good, 30%+ strong, 40%+ elite | 40 to 50% good |
None of the three gives a sample size, and Exposure calls its own bands directional.
Exposure and Hawky both report lower hook rates in Reels than in Feed. Hawky puts the gap at 5 to 10 percentage points. Hook rate on reach also reads higher than on impressions, per Exposure, so check which base a benchmark used.
Compare Each Video With Your Account Average
Even the publishers of those ranges point back to your own numbers. Exposure tells you to judge an ad against your account history first. Hawky says your historical median is often the most useful benchmark.
Our free Video Hook Rate Analyzer skill works the same way. It compares each video with your account's video average first, using the export you upload to your own Claude or ChatGPT chat. It never logs in to your ad account or changes anything, and nothing is sent to Agency AI.
Its rule flags a hook as weak only below 85 percent of your account average and below 25 percent. Hold rate gets the same test against 40 percent.
So if your videos average a 20 percent hook, the line drops to 17 percent. Then a 22 percent hook sits above your norm.
Hook Rate vs Hold Rate: How to Read Them Together
Read hook rate, hold rate and link click-through rate together, and each weak spot points to its own edit.
Take a hypothetical Shopify store with four prospecting videos over 30 days.
| Video | Impressions | 3-Second Plays | ThruPlays | Hook Rate | Hold Rate | Link CTR |
|---|---|---|---|---|---|---|
| Founder intro, 0:45 | 36,000 | 5,760 | 2,880 | 16.0% | 50.0% | 0.9% |
| UGC review, 0:30 | 30,000 | 10,500 | 2,100 | 35.0% | 20.0% | 1.0% |
| Product demo, 0:25 | 24,000 | 7,440 | 3,348 | 31.0% | 45.0% | 0.5% |
| Before and after, 0:20 | 20,000 | 6,600 | 3,300 | 33.0% | 50.0% | 1.5% |
| Account total | 110,000 | 30,300 | 11,628 | 27.5% | 38.4% | 0.95% |
This account's weak-hook line is 23.4 percent, 85 percent of its 27.5 percent average.
Weak Hook: Replace the Opening Only
Just 16 percent of the founder intro's impressions become 3-second plays. Yet half of the people it does stop stay for a ThruPlay, so the body works. Keep the middle and the ending, and cut or shoot a new first 3 seconds.
Good Hook, Weak Hold: Find Where Viewers Leave
At 35 percent, the UGC review has the best hook in the account. Only 20 percent of those viewers reach a ThruPlay, well under the 32.6 percent hold line. Its drop-off curve shows where they go.
Divide each of the video plays at 25, 50, 75 and 100 percent counts by 3-second plays. Those counts include viewers who skipped ahead, and the shares form your video retention rate curve.
On this 30-second video, only 40 percent of 3-second viewers reach the 25 percent mark, 7.5 seconds in. That fall, the steepest on its curve, suggests an opening that promises more than the middle delivers. Cut the intro and show the product sooner.
Good Watch, Few Clicks: Fix the Ask
With a 31 percent hook and a 45 percent hold, the product demo keeps viewers watching. Its link click-through rate is only 0.5 percent, against 0.95 percent for the account. Add the offer on screen, say the call to action out loud, and close on an end card with the offer.
Before and after beats the account average on all three numbers, so it is the winner. Extend its life with new hooks. When you scale Facebook ads on Shopify, keep each budget step small enough that learning does not restart.
When all three numbers look healthy and sales still lag, the video has done its job. Ads often land on a Shopify product page, so check that page, your price and the offer before you touch the video.
How to Improve Hook Rate on a Video Ad
Rework the first 3 seconds and leave the rest alone. Meta's research with Nielsen, shared in 2016, found up to 47 percent of a video campaign's value came in its first three seconds.
Its advice for mobile video ads is to capture attention fast and lead with your most captivating elements. It also covers sound-off viewing, captions, taller 4:5 or 9:16 frames, and a message in 15 seconds or less.
To improve your hook rate, pick from the skill's checklist of strong openings:
- Put motion or a person in the first frame, with no logo intro.
- Add bold on-screen text in the first second: the problem, the result or a surprising claim.
- Show the product in use, or the result, inside 2 seconds.
- Break the pattern with an unexpected angle, a close-up or a question to camera.
- Shoot for a native Reels look: on a phone, vertical 9:16, captions on.
Test New Hooks on the Same Body
Run three new hooks on the same body, inside the same ad set as the original. That way the test reads only the opening. Adding an ad can restart that ad set's learning, though. On your biggest ad set, use the creative test from our guide to Facebook creative testing. Judge them after 7 days on hook rate first, then on cost per purchase.
Change the visuals as well as the words. A swapped headline on the same image gives Meta Andromeda little new to work with.
Winners wear out too. Creative fatigue on Meta sets in when an audience has seen the same creative too many times.
A new hook gives a proven body a fresh first impression. It also keeps your weekly creative pipeline full, step 4 in our plan to automate Facebook ads for Shopify.
How Agency AI Helps You Rebuild Video Ads
Agency AI, our Shopify app for Meta and Google ads, helps you launch the edited version. In Create Ad, you pick Video as the ad type, then New Customers or Remarketing as the audience.
You can start a new ad or duplicate an existing one. Each copy field has Generate with AI, with 12 copywriting angles. Building a full Meta ad takes about 2 to 3 minutes.
After launch, our 90+ rule engine reads the last 7 days of data and measures each recommendation against your break-even ROAS. Nothing changes until you click Apply.
Through our MCP connector, Claude or ChatGPT can read your live Meta, Google and Shopify data. Any write it proposes runs only after you confirm it.
Final Thoughts
Hook rate grades one job: stopping the scroll. Hold rate grades the next one, and link clicks grade the last.
Read all three against your own account average, then change only the part that failed. A weak opening needs a new opening, not a reshoot.
Frequently Asked Questions
Sources
Meta's Business Help Center supplies the metric definitions. Two Meta for Business articles supply the video advice and the Nielsen figure.
Benchmark ranges come from blog posts by Segwise (March 23, 2026), Hawky (March 24, 2026) and Exposure (June 16, 2026).
The four-video store is hypothetical. Skill rules and Agency AI features come from our own documentation.
Free to install. Launch campaigns and pay nothing for your first 30 days.
