Measurement library
What unmeasured should mean in an AI visibility report
Not cited means a check completed, returned a usable source list, and your domain was not in it. Unmeasured means the check never produced usable evidence, so it says nothing either way. A report that renders both as a zero is asserting a negative result it did not observe, and the error is systematic rather than random: provider outages can only ever push the reported score down.
Published 2026-08-03. Reviewed 2026-08-03. Published by ShiftNode Digital s.r.o.
The distinction
The test is simple: can you open the recorded answer and the returned source list for that check? If yes, a missing domain is a real negative. If no, there is nothing to conclude.
- Not cited is an observation. Evidence exists, it was inspected, and the domain was absent.
- Unmeasured is the absence of an observation. No evidence exists to inspect.
Why merging them is not a rounding error
If unmeasured checks became zeros, the reported score would fall whenever a provider had a bad day, and it would fall furthest for the services that are least reliable. Two audits of an unchanged site would disagree, and the disagreement would be read as a visibility change.
It also breaks comparison across plans. A plan that checks five services has more opportunities to hit an outage than a plan that checks one, so counting failures as misses would systematically penalise broader coverage.
What a report should show instead
- The completion rate for every service, next to its result.
- An explicit unmeasured state in the interface, visually distinct from a zero.
- The recorded answer and source list behind every non-failed check, so a negative can be inspected.
- A refusal to publish an aggregate when completion falls below a declared threshold.
What this page does not establish
- This is a reporting rule, not a ranking factor. Handling unmeasured checks correctly makes a metric trustworthy; it does not make a site more visible.
Sources
Run a free audit to see this measurement format applied to your own domain, or read the full measurement method.