Glossary
Hold rate: how many people stay to the end of the video
Hold rate shows how many of the people who started your video watched it through. The formula, what a fair number is, and how it connects to hook rate.
Also available in: Português · Español
Hold rate is the share of people who watched the whole video out of those who started watching it. If hook rate measures the door, hold rate measures the walk. It answers whether your argument reached its end or was abandoned halfway.
Formula
Hold rate = plays to 100% ÷ 3-second plays × 100.
A video with 3,600 three-second plays and 540 complete plays has a 15% hold rate.
The denominator is the decision that matters here. Using impressions instead of 3-second plays folds two questions — who stopped and who stayed — into one number, and you lose the ability to know which one broke.
What an acceptable number looks like
It depends on length, and there is no way around that. A 15-second video at 25% and a 90-second video at 8% can be delivering the same commercial outcome. Comparing them head to head tells you nothing.
The comparison that works is between ads of similar length and role, in the same campaign and the same period. Inside that frame, hold rate ranks creatives precisely.
| What you want to know | Where to look |
|---|---|
| Does the opening hold? | Hook rate |
| Does the argument survive the middle? | Hold rate |
| Did anyone act afterwards? | CTR and cost per result |
What makes the number drop
- The opening's promise is not kept. High hook with low hold is the classic bait pattern: it grabbed attention and did not deliver.
- A flat middle. The drop usually concentrates at one specific point. Watch your own video looking for where it slows down.
- Runtime longer than the content deserves. Thirty seconds of argument stretched to seventy loses people to boredom, not disinterest.
- An audience too cold for the format. Long video works better on people who already know the brand. In cold audiences, patience is shorter.
How to use it
Hold rate decides editing, not campaigns. The correct use is diagnostic: when cost per result rises and hook rate stays flat, hold is usually what fell — people still stop, they just do not stay for the offer.
The fix is reordering the video, not reshooting it. Moving the main argument ahead of the drop-off point solves more cases than producing a new piece. And if you run the video objective, it helps to understand the bar the platform uses in ThruPlay.
To bring hook, hold and result into one reading, the template is on the paid media reporting page.
Frequently asked questions
What is the hold rate formula?
Plays to 100% divided by 3-second plays, times 100. The denominator is everyone who started watching, not everyone the ad was shown to.
Does a low hold rate mean a bad video?
Not always. Longer videos have naturally lower hold rates, and that is not a defect. What matters is whether the essential message lands before the point where people leave.
Do hold rate and ThruPlay measure the same thing?
No. ThruPlay is an absolute count of qualifying plays; hold rate is a ratio between who finished and who started.