Hold rate: are people watching past the hook?

How Meta defines the parts

Hold rate itself isn't a Meta metric. Like hook rate, it's a ratio advertisers build from these counts.

Sources: [1] [2]

Two ways to work it out

• ThruPlays ÷ 3-second video plays: the most common version. Of the people the hook caught, how many stayed to 15 seconds or the end.
• ThruPlays ÷ impressions: some advertisers use this instead. It mixes hook and hold into one number, so it's lower and harder to read.

Build it as a custom metric in Ads Manager the same way as hook rate (Columns, Customize columns, Create custom metric). Write the formula in the description so whoever reads the report knows which version it is.

Why ad length changes the number

Because a ThruPlay is 15 seconds or the end, whichever comes first, a 10-second ad only scores a ThruPlay on a full watch, while a 30-second ad scores one halfway through. Comparing hold rate across ads of very different lengths isn't fair. Here's where 15 seconds falls in our own formats:

Sources: [1]

What a low hold rate usually means

If hook rate is fine but hold rate drops, the hook made a promise the next scene didn't keep, or the middle repeats itself. Look at what happens between seconds 3 and 8. In our problem-solution ad, that's the problem scene and the first look at the product, and it's where we'd change things first.

Questions

What is a ThruPlay on Facebook?

Meta counts a ThruPlay when your video plays to the end or for at least 15 seconds, whichever comes first. It's also an optimization goal you can pick for video ads.

What's the difference between hook rate and hold rate?

Hook rate is 3-second plays divided by impressions: did the opening stop people. Hold rate is usually ThruPlays divided by 3-second plays: did the people who stopped keep watching.

Is a ThruPlay a completed view?

Only for videos 15 seconds or shorter. For longer videos it counts at 15 seconds, so someone can skip the end and still count as a ThruPlay.