Proof, with its scope declared: how to read an AI visibility number

Every vendor in AI visibility shows you a chart going up. Almost none will tell you what the chart is made of. We decided to be the exception — not because it's noble, but because in this category, undeclared numbers are how clients get burned.

The uncomfortable truths about AI visibility measurement

Different tools measure different things and call them the same name. One platform reads an engine's consumer interface; another queries its API; a third simulates the retrieval. They can return different answers to the same question on the same day — and every one of them will label the result "visibility." A before/after built from two different instruments isn't a trend. It's two unrelated photographs.

One scan is one scan. AI answers vary run to run. A brand can appear at #1 on Tuesday and be absent Thursday — same prompt, same engine. Which means a screenshot is a fact about a moment, not a claim about a position. We hold results to a sustain standard before we call anything won: the same reading, weeks in a row, same instrument, same wording.

The wording is the measurement. Change three words in a prompt and you've changed the question — and often the answer. "Best physiotherapy clinic near me" and "best WSIB physiotherapy clinic in Brampton" surface different businesses for different reasons. Any comparison that doesn't hold prompt wording identical, character for character, is comparing apples to a photograph of oranges.

So this is what we declare, on everything

Which prompts, exactly. Which engines. Which market. Which instrument took the reading, and when. And whether the after was re-scored on the identical wording as the before.

You'll see it on every case study we publish — a scope line stating all five, including, where it matters, which readings we excluded because the instruments weren't comparable. When our own measurement changed tools mid-engagement, the case study says so rather than splicing the trend line. That entry is weaker-looking for it and more true — a trade we'll take every time.

Why this is the buying criterion

If you're evaluating anyone in this space — us included — the test takes one question: "Same prompts, same engine, same tool — show me the before and after." A provider measuring honestly answers instantly, because the discipline is already in their workflow. A provider who hesitates is showing you two photographs.

Numbers with their scope declared are slower to produce and harder to inflate. That's precisely the point — and precisely what your renewal decision deserves to stand on.

See what a scope-declared reading of your own brand looks like. → Get your Trust Index