I looked at an account last week where the team had killed an ad with a 28% hook rate. It was generating a $30 CPA at $5K/week in spend. The ad they replaced it with had a 44% hook rate. It hasn't cleared $80 CPA in two weeks.
───
The Number Everyone Watches, and What They're Actually Watching For
Hook rate -- the percentage of viewers who watch the first 3 seconds of your ad -- has become the de facto first-pass metric in DTC creative. Open almost any performance review and it's in the top row. "What's the hook rate?" gets asked before "what's the CPA?"
I get it. It's immediate, it's clean, and it feels like a signal about creative quality. If people aren't stopping, the ad didn't work.
Except that framing gets it backwards. Hook rate tells you whether the first 3 seconds worked. It doesn't tell you whether the ad worked. Those are different questions, and collapsing them is how good creative gets killed.
───
Context Changes Everything
A 45% hook rate on an ad running at $200/week is almost meaningless. At that spend level, you don't have enough signal to trust any metric -- not hook, not hold, not CTR. The algorithm is still learning. Your audience pool is small. You're reading noise.
A 28% hook rate on an ad running at $5K/week with a $30 CPA is an ad worth studying. That ad is working at scale. The first 3 seconds are getting 28 out of every 100 people -- and those 28 people are converting.
The mistake is treating hook rate as an absolute score instead of what it actually is: a relative input. A high hook rate on a dead ad is useless. A lower hook rate on a converting ad is a data point to understand, not a failure to fix.
───
What the Combinations Actually Tell You
Here's the diagnostic framework I use when I'm reading a creative:
High hook + low hold rate: The first 3 seconds overpromised. The headline was clickbait. The visual grabbed attention that the body couldn't earn. Fix is in the body, not the hook.
High hook + high hold + low CTR: The ad is genuinely engaging but it's not creating urgency or a clear reason to act. People are watching but not clicking. This is a copy/CTA problem, not a creative problem.
High hook + high hold + high CTR + high CPA: The ad is doing everything right until the landing page. Stop pulling the creative and go fix the page.
Lower hook + strong CTR + strong CPA: This is the ad I want to understand. Why are the people who stay converting at that rate? What is it about this ad that's selecting for the right buyer? That's worth more to me than a high hook rate on an ad that attracts everyone and converts nobody.
───
The Question You Should Be Asking
Most creative reviews start with: "What's the hook rate?"
The better question is: "What is my hook rate telling me about what broke next?"
Hook rate is the first chapter. It tells you whether people opened the book. It doesn't tell you if the story was any good.
When I'm diagnosing creative performance, I'm reading all four metrics together -- hook, hold, CTR, CPA -- and I'm looking for where the chain breaks. The break is the signal. Hook rate just tells you whether the break happened before 3 seconds or after.
───
What to Actually Do With This
Pull your last 30 days of creative. For any ad you paused or killed in the last two weeks, check whether you had meaningful spend data -- at least $1,000 -- before you made that call. If you didn't, you might have cut an ad based on noise.
Then look at your current top performers by spend. Note where each one's hook rate falls. Now ask: is the hook rate the reason it's working, or is something downstream doing the heavy lifting?
The answer will change how you brief the next batch.
───
Takeaway
Before you kill an ad for a low hook rate, pull the hold rate, CTR, and CPA -- if two of those three are strong, you have an audience problem, not a creative problem.
───
One More Thing
There's a pattern I keep seeing where teams set a hook rate floor -- "anything below 30% gets cut" -- and enforce it regardless of what the other metrics say. If you have a rule like that in your creative review process, it's worth revisiting. Rules are useful when they save you from noise. They're harmful when they override signal. A $30 CPA is signal.
-- Sadie