ROAS tells you an ad isn't working. It never tells you why. That's the actual problem with judging creative by one blended number, because a failing ad can die at three completely different moments, and each one needs a totally different fix.

We track three specific metrics for exactly this reason, and together they answer three separate questions. Are you earning attention. Are you holding it. Are you converting it into action.

Thumbstop ratio answers the first question

This is simply how many people stop scrolling when your ad shows up. It's the earliest possible signal of relevance, and if an ad can't earn that first half second of attention, nothing downstream matters at all. A weak thumbstop ratio isn't a targeting problem or a budget problem, it's usually a visual problem, the first frame isn't doing its job.

Hook score answers the second question

Stopping the scroll and holding attention are two different skills. Hook score tells you whether the opening seconds actually keep someone watching once they've stopped, which tells you whether the angle and message are landing or whether the concept falls apart the moment someone gives it a real second of thought. A strong thumbstop with a weak hook score usually means the visual promised something the opening lines didn't deliver.

Click to purchase rate answers the third question

This measures how effectively the people who click actually turn into buyers. It's the metric that separates real purchase intent from simple curiosity, and when it's low, the honest answer is usually not spend, it's audience mismatch. You're attracting the wrong person efficiently, and no amount of extra budget fixes that, because you'd just be paying to attract more of the wrong person faster.

Why this actually matters for how you fix things

Match the fix to the exact metric that's weak, not to the ad as a whole. A scroll problem needs a new opener, a static thumbnail swap, a different first frame, not a whole new script. A hold problem needs a new angle or a sharper message, since the concept itself isn't landing. A conversion problem needs a new audience entirely, since the creative might be doing its job perfectly and just reaching the wrong person.

Pull these three metrics for your current top performing ads this week and map each weak number to its actual stage. You'll stop guessing at what to fix, and start fixing the specific thing that's actually broken.