Evidence-first writing on AI search, and what the GEO market gets wrong.
I publish the methodology, the sources, and the results I can actually reproduce. When the evidence says a popular tactic does nothing, that gets published too.
The local AI visibility study protocol, published first
I have not collected a single data point for this study. This page is the entire method, fixed on 29 July 2026, published before capture opens so that anyone can run it a
Evidence Audits and Primary SourcesThe AI Overviews click studies do not agree
Ten published measurements, an 84 point spread, and not one of them is wrong. The disagreement dissolves the moment you sort them by what they actually measured.
The Cited Method (ACCESS, MEASURE, MAP, EARN, PROVE)How many prompts before the number means anything
Every measurement post asserts a prompt count. None of them show the arithmetic. Here is the arithmetic, run three times, for three different claims.
Evidence Audits and Primary SourcesDoes llms.txt Do Anything? I Checked the Logs
Both sides of this argument have been running on first principles for two years. Three server log studies, one correlation study and one live audit later, the argument is
The Cited Method (ACCESS, MEASURE, MAP, EARN, PROVE)Can Every AI Crawler Actually Fetch This Site?
Reading robots.txt tells you what a site intends. It tells you nothing about what an AI agent actually receives. This is the procedure for the four layers underneath, and
Evidence Audits and Primary SourcesWhat Google's Docs Say vs What GEO Invoices Say
Google's public documentation has a section called "Mythbusting generative AI search: what you don't need to do." It names five things. I fetched five live GEO service pa
Evidence Audits and Primary Sources
The local AI visibility study protocol, published first
I have not collected a single data point for this study. This page is the entire method, fixed on 29 July 2026, published before capture opens so that anyone can run it a
Evidence Audits and Primary SourcesThe AI Overviews click studies do not agree
Ten published measurements, an 84 point spread, and not one of them is wrong. The disagreement dissolves the moment you sort them by what they actually measured.
Evidence Audits and Primary SourcesDoes llms.txt Do Anything? I Checked the Logs
Both sides of this argument have been running on first principles for two years. Three server log studies, one correlation study and one live audit later, the argument is
Evidence Audits and Primary SourcesWhat Google's Docs Say vs What GEO Invoices Say
Google's public documentation has a section called "Mythbusting generative AI search: what you don't need to do." It names five things. I fetched five live GEO service pa
The Cited Method (ACCESS, MEASURE, MAP, EARN, PROVE)
How many prompts before the number means anything
Every measurement post asserts a prompt count. None of them show the arithmetic. Here is the arithmetic, run three times, for three different claims.
The Cited Method (ACCESS, MEASURE, MAP, EARN, PROVE)Can Every AI Crawler Actually Fetch This Site?
Reading robots.txt tells you what a site intends. It tells you nothing about what an AI agent actually receives. This is the procedure for the four layers underneath, and