Sample report · run on ourselves · 26 August 2026
We sell AI visibility tracking. AI does not mention us.
This is a real report, run on our own domain, published unedited. We picked our own category because a sample where the subject looks good tells you nothing about what happens when they do not.
The scoreboard
- Questions asked
- 6
- Citations recorded
- 176
- seobuilder.tech cited
- 0
- Engines asked
- 7
176 citations went to other domains across the six questions our own buyers ask. None of them was ours. Four competitors were named: tryprofound.com, otterly.ai, semrush.com and seranking.com.
Question by question
what is the best ai visibility tracking tool
6 of 7 engines answered · 28 domains cited · we were not one of them
how do i track whether chatgpt mentions my brand
6 of 7 engines answered · 21 domains cited · we were not one of them
Competitors cited here: otterly.ai, semrush.com
what are the best geo tools for generative engine optimization
7 of 7 engines answered · 27 domains cited · we were not one of them
Competitors cited here: tryprofound.com, semrush.com
what is the cheapest way to monitor brand mentions in ai answers
6 of 7 engines answered · 26 domains cited · we were not one of them
is an ai visibility tool worth paying for
6 of 7 engines answered · 19 domains cited · we were not one of them
Competitors cited here: tryprofound.com
what is the best rank tracker for a small agency
7 of 7 engines answered · 22 domains cited · we were not one of them
Competitors cited here: seranking.com
The gap table
In the paid reportEvery question you lose, the exact URL that won it, its rank among cited sources, and which engines it spans. This is the part people paste into a ticket tracker.
What kind of page is winning
In the paid reportEach winning URL classified by type: third-party comparison, vendor docs, forum thread, video, listicle. It tells you whether the gap is a page you can write or one someone else has to write about you.
The next 30 days
In the paid reportFive to eight items, each naming the question, the engine, the exact URL and what to do. No effort scores and no impact scores, because we cannot measure impact and inventing a number is worse than leaving it out.
What this sample does not tell you
- We cannot tell you why we are absent. We tested the usual explanations on our own corpus and they did not hold: schema, list counts and heading counts did not separate cited pages from uncited ones once page length was controlled.
- Engines are not deterministic. Ask the same question tomorrow and the cited sources can differ. This is one run.
- Six questions, one category, US English. A real report asks 20 and includes three competitors you choose.
- A citation is not a visit. We measured whether an engine named a domain, not whether anyone clicked.
Every number above is checkable at /ai-citations, which is why we ran it on ourselves.
Your version asks 20 questions instead of 6, against three competitors you name, with the three locked sections filled in by a person.
Start a report · $99