What an AI visibility report should contain
A useful AI visibility report shows five things: the exact questions asked, which assistants were queried, every company named in each answer, the sources those answers cited, and the date it ran. Anything less is a score you cannot check, cannot reproduce, and cannot use to judge whether next month was better.
Why does the shape matter more than the score?
A number on its own cannot be argued with, which sounds like a strength and is the opposite. If a report says you scored 42, you cannot tell whether that is good, what would move it, or whether the same test next month is even the same test.
What are the five things?
| What | Why it matters |
|---|---|
| The exact questions | Different phrasings return different companies. Questions chosen after the fact can be chosen to flatter. |
| Which assistants | Answers differ by engine. A report on one engine is one engine's opinion. |
| Every company named | This is the competitive picture, and it is often not who you assumed. |
| The sources cited | It tells you which pages the answer was assembled from, which is where the work goes. |
| The date | Answers move. A reading with no date cannot be compared to anything. |
What does a single run actually prove?
Less than people think. Answers vary between runs, so one reading is a sample rather than a verdict. The value appears when the same questions are asked again on a schedule and the drift is visible.
What should make you suspicious?
A score with no working. A refusal to name the engines. Questions that read like keywords rather than how a customer speaks. And any date attached to a prediction, because assistants read from indexes that refresh on their own schedule and nobody doing this work controls that calendar.
What our free report shows
Five buying questions, put through a live web search one question at a time, with the companies that came back for each and whether yours was among them. No call and no card. The method we use on our own business, including what we do not record, is published on our methodology page.
Questions
What should an AI visibility report contain?
The exact questions asked, which assistants were queried, every company named in each answer, the sources those answers cited, and the date it ran. A single score with no working shown cannot be checked or reproduced.
Is one report enough to know where I stand?
No. Answers vary between runs, so a single reading is a sample rather than a verdict. The useful signal comes from asking the same questions again on a schedule.
How much should an AI visibility report cost?
Ours costs nothing and needs no call. What is worth paying for is the work that follows it and the monthly re-measurement, not the first reading.