The Setup
Say you’ve got two ads running to the same audience with the same offer. One is at a 4 ROAS and the other is at 2.
The obvious move is to switch off the 2 and put the money behind the 4. Sometimes that’s right. But before I do anything, I check one number first, and it isn’t ROAS.
The Number I Check First
Hook rate is the share of people who watched past the first 3 seconds of your video. On Meta it’s usually worked out as 3-second plays divided by impressions. It tells you whether the opening of the ad earned any attention at all.
I look at it first because it splits the problem in two. Either people aren’t staying, or they’re staying and not buying. Those need completely different fixes, and ROAS alone can’t tell you which one you’ve got.
How I Read The Two Together
The row I care about most is high hook rate with low ROAS. It’s the one that gets misread. The instinct is to say the ad has failed and replace it. But the ad has done its job: it got attention and held it. The leak is somewhere after that.
| Hook Rate | ROAS | What It Usually Means | Where I’d Look |
|---|---|---|---|
| High | High | The ad is working | Protect it, then plan what comes next as it fatigues |
| High | Low | People watch but don’t buy | Landing page, offer, price, or the wrong audience |
| Low | High | Few people stay, but the ones who do convert | Test a stronger opening, there’s probably more here |
| Low | Low | Nobody is staying | The first 3 seconds, before anything else |
When A New Hook Won’t Help
If hook rate is already high and ROAS is still low, a new opening line is the wrong fix. Nobody is leaving in the first 3 seconds. They’re leaving later, or after the click.
That’s usually the landing page, the offer, or a mismatch between what the ad promised and what the page delivers. Look at the page on a phone, because that’s where most of your traffic is. Check that the product in the ad is the product people land on. Check the price isn’t a surprise.
When Nothing After The First 3 Seconds Matters
If hook rate is low, it doesn’t matter how good the middle of the ad is. Most people never saw it.
In that case the first 3 seconds are the whole job. A different first frame, a stronger first line, or a different person on screen can change the result without touching anything else in the ad. I’d test openings before I’d touch the rest.
This is also why a high CTR can mislead you. In one account I worked on, the ad with the highest click-through rate converted worse than an ad where the founder just talked to camera. Clicks measure curiosity. They don’t measure how many of those people were the right ones. (More on that here.)
A Few Cautions
- Wait for enough data. A 4 against a 2 after one day and a handful of purchases can flip by the weekend. Don’t call a winner on a thin sample.
- Compare like with like. Same audience, same offer, similar spend. Otherwise the comparison tells you nothing.
- Hook rate isn’t a goal on its own. A funny opening that attracts the wrong people will post a great hook rate and a bad ROAS.
What This Means For Your Account
Next time you’re about to switch off an ad, pull its hook rate first. If people aren’t staying, work on the opening. If they are, look at the page. You’ll fix the right thing, and you’ll stop killing ads that were only one fix away from working.
For the fuller framework on what’s worth tracking and what isn’t, see The Paid Social Metrics That Actually Matter (And The Ones You Can Ignore.)