99.9% earned the click.
The page changed the subject.
Orca has a rare asset: “99.9% of fixable AI vulnerabilities remain unpatched.” It is specific, it creates urgency, and it gives a security leader a clear reason to care. But the ad leads with a broad headline, and the landing page resets the conversation with a report title. The funnel does not lack proof. It delays it.
The ad and the page it lands on
What the roast found · tap to open each one
The dark design and the oversized number make the risk impossible to miss. Attention is not the problem here. What happens around that attention is.
“The Full Picture of AI Risk” could point to AI-generated code, data exposure, agent security, Shadow AI, governance or model risk. The stat carries the creative while the headline stays broad.
Reverse it: “99.9% of Fixable AI Vulnerabilities Are Still Unpatched.” Build the ad around what stopped the scroll.
A CISO who sees 99.9% immediately asks 99.9% of what, across what sample, and how it was derived. The ad does not answer.
Add the proof Orca already has: “Based on real-world telemetry from 1,200+ production organizations.”
“2026 State of AI Security Report” is accurate, but it resets the conversation. The page answers much of what the buyer wants to know, only later.
Carry the same claim above the fold, then support it with what the telemetry reveals.
The creative says “Get the Full Report,” the button says “Download,” and the page says “Get the Report.”
One action, one promise. Then move one more stat next to the form, because big claims need immediate backing.
Why it scores 7/10 · tap a row
Questions, answered
This teardown is based on a real AdRoast roast of Orca Security’s live ad. See the full roast →
More teardowns
Your ads, roasted the same way
Pull your live ads, score each against your buyer, see where the click leaks.