Blog
What is the difference between thumb-stop rate and hold rate?
What is the difference between thumb-stop rate and hold rate?
The difference between thumb-stop rate and hold rate is two jobs. Thumb-stop is colloquial for whether the opening interrupts scroll. On Meta it maps to hook rate: 3-second video plays divided by impressions. Hold rate is whether people who started stay: ThruPlays divided by 3-second plays; Optimize them separately; One metric cannot diagnose both.
What is the difference between thumb-stop rate and hold rate?
The difference between thumb-stop rate and hold rate is two jobs. Thumb-stop is colloquial for whether the opening interrupts scroll. On Meta it maps to hook rate: 3-second video plays divided by impressions. Hold rate is whether people who started stay: ThruPlays divided by 3-second plays; Optimize them separately; One metric cannot diagnose both.
Two jobs, two ratios
Thumb-stop / hook rate asks: did frame one earn a start? Hold rate asks: did the body keep the people who started? Meta documents the underlying plays; the ratios are operator math.
Raising thumb-stop with clickbait that the body cannot pay off tanks hold. Raising hold with a slow, honest intro can miss the feed entirely. Read both before you rewrite.
How to work each lever
For thumb-stop: change only the opening on a locked body. Visual anchor, text, motion, face. For hold: change payoff timing, proof, and pacing after second three. Do not mix both in one sandbox batch.
There is no official “healthy thumb-stop” or “strong hold.” Compare variant A to variant B in the same window. Internet league tables are not Ads Manager.
What to make next
Diagnose which ratio slipped, then isolate that layer. eonik assembles the on-brand opening or body rewrite. You approve. You still read both ratios in Ads Manager.
Do not optimize “the video” as one knob. Thumb-stop and hold fail for different reasons.
Thumb-stop vs hold
| Ratio | Formula (operator) | What to change |
|---|---|---|
| Thumb-stop / hook rate | 3-second plays ÷ impressions | Opening only |
| Hold rate | ThruPlays ÷ 3-second plays | Body payoff and pacing |
| Both together | Not a single Meta column | Diagnose first; isolate second |
Fix thumb-stop first when
- Impressions are plentiful but 3-second plays are weak vs last week
- Frequency is rising on one opening
- The body still holds on remaining viewers
Fix hold first when
- 3-second plays look fine vs baseline
- ThruPlay / completion falls after the hook
- Viewers bounce right after the promise
What usually breaks
- Treating thumb-stop and hold as one score.
- Clickbait openings that the body cannot pay off.
- Rewriting hook and body in the same sandbox batch.
When the readout is fatigue, hook rate, or velocity, the next move is consider what to make. Then make the cut in eonik and approve it. The loop is how to generate AI ads. It never touches spend.
“Almost any question can be answered, cheaply, quickly, and finally, by a test campaign. And that’s the way to answer them, not by arguments around a table.”
That principle still holds. The next on-brand cut is yours to approve.
Sources
Questions
Is thumb-stop the same as hook rate?
In operator language, yes. Both describe 3-second video plays divided by impressions. Meta does not name “thumb-stop rate” as an official metric. It documents 3-second video plays; You do the division; Use the same placement and window when you compare variants, and do not import a blog’s invented healthy thumb-stop number into your stop rules.
What is a good hold rate?
Meta does not publish one. Hold rate as operators use it is ThruPlays (completion or at least 15 seconds) divided by 3-second plays. Compare variant A to variant B, or this week to last week, on the same account. There is no universal good hold rate; Label which numerator you chose so the readout stays honest.
What is the difference between thumb-stop rate and hold rate?
Thumb-stop / hook rate measures whether the opening earned a start. Hold rate measures whether people who started stayed through ThruPlay or your 15-second proxy. They fail independently: a loud opening can raise thumb-stop and destroy hold, and a slow honest body can hold viewers who never stopped. Read both, then change only the layer that slipped.
Free, no card · macOS 15+ · Apple silicon