Why we do / don’t

Why is a failed measurement reported as unknown instead of zero?

A failed measurement is reported as unknown instead of zero because zero is a finding and unknown is the absence of one. Reporting a collection failure as zero would turn our pipeline problem into an apparent visibility drop, and you would spend the week auditing your pages instead of asking us.

Zero says an engine answered and did not mention you. Unknown says we did not get a usable answer. Those are different facts and they lead to different actions.

Collapsing them is the flattering direction only until someone checks. A chart that silently converts outages into drops looks responsive and teaches you to distrust it the first time you trace one.

The same principle governs crawler verification, where an unverifiable visit is its own reported state rather than being folded into verified or spoofed.

What this answer is based on

Unknown-state reporting policy — a failed run is never a zero