AI visibility case study: how RaceX moved from 63 to 94
A transparent remediation example from one of our own ventures, including what changed and what the score still did not prove.
Read: AI visibility case study: how RaceX moved from 63 to 94 →
Notes
AI News
Notes on how AI systems interpret, trust and recommend businesses.
Many sources go in. One answer comes out — and it is not the same answer twice.
AI search citations: mentions, sources and recommendations are not the same thing
A cited page, a named brand and a recommended provider are three different outcomes. Measure them separately.
Read: AI search citations: mentions, sources and recommendations are not the same thing →
Brand mentions and AI visibility: what the 75,000-brand data actually says
The correlation is strong, the caveat matters, and “mentions are the new backlinks” is still too simplistic.
Read: Brand mentions and AI visibility: what the 75,000-brand data actually says →
AI visibility vs GEO vs AEO: what the terms actually mean
GEO and AEO describe methods. AI visibility is the outcome you can actually measure.
Read: AI visibility vs GEO vs AEO: what the terms actually mean →
AI crawler access: robots.txt, OAI-SearchBot and JavaScript
Crawler access is an eligibility gate. Different assistants use different agents, and blocking one can have a different effect from blocking another.
Read: AI crawler access: robots.txt, OAI-SearchBot and JavaScript →
Entity SEO for AI search: clarity before cleverness
Structured data can make relationships explicit. It cannot manufacture authority or guarantee citations.
How AI assistants choose which businesses to recommend
Think in four stages: eligibility, retrieval, confidence and selection — then measure the output.
Read: How AI assistants choose which businesses to recommend →
Does ChatGPT know my business? How to test it properly
Knowing your name and recommending you to a buyer are two different tests.
Read: Does ChatGPT know my business? How to test it properly →
AI visibility checker: what a good one should actually test
A score is only useful if you can see the evidence, the method and what the tool did not measure.
Read: AI visibility checker: what a good one should actually test →
How to improve AI visibility: the practical order
Five steps, in order: access, identity, useful answers, corroboration and repeated measurement.
What is AI search visibility?
Whether AI assistants understand your business well enough to recommend it. It is not a ranking, and it works differently to search.
AI trust signals, explained
Which signals are supported by evidence, which are repeated without any, and what actually moves the needle.