Skip to content
wrath
WHAT WE MEASURE

Seven readings
of one answer.

Each dimension asks a different question of the same material. Together they describe whether a brand is found, trusted, readable and accurately represented.

Every finding carries its answer.

A finding is not a judgement written after the fact. It is bound to the engine response, page or source it traces back to, and the wording is kept so it can be read again.

A citation is not a cause.

A displayed source shows that the engine attached it to a nearby claim. It does not show that the source was retrieved, that it shaped the wording, or that it decided the recommendation. Those remain inferences, and they are labelled as inferences.

There is no permanent position.

Answers move by engine, by question, by date and by context. One answer is an observation. A baseline is what a set of them looks like when the questions are held still, which is why the same questions are reused every time.

What the measurement will not do.

  • Where evidence is thin, a component is marked not computable. Never scored zero — measured-and-bad and not-measurable are different findings.
  • Where a reading is structurally unsound, it halts and says why rather than publishing a number that would mislead.
  • Where two readings disagree beyond tolerance, the disagreement stays visible rather than being averaged into a cleaner answer.

Where to start.

The free demo gives a focused, fast view of AI visibility and page readiness where there is enough data, showing the largest observed leaks and the evidence behind each. The full engagement expands coverage across all seven dimensions, then continues into fixes and re-measurement.

Get a free demo

Find out what the answer says about you.