Field Research
Research
notebook
Field studies, calibration data, and methodology notes from running AEO audits on real businesses. Published in public, including the parts where we discovered our own work was wrong.
Entries
02
Latest
2026-04-25
Method
Live signal-checks against
ChatGPT · Perplexity ·
Gemini · Claude
ChatGPT · Perplexity ·
Gemini · Claude
N°012026-04-25
Field Study6 min read
We Ran Our AI Visibility Audit On Eight Famous Brands. Only One Passed.
Wikipedia, Apple, Microsoft, OpenAI, Anthropic, Perplexity, Facebook, and Google walked into our scoring engine. Seven walked out with grades that surprised us — and one bug we found in public.
Read
N°022026-04-25
Methodology Note4 min read
ChatGPT Won't Recommend Anthropic. Claude Won't Recommend OpenAI. We Have the Data.
A 4x4 matrix of LLM raters scoring LLM creators surfaced a reciprocal suppression pattern between OpenAI and Anthropic specifically. The pattern doesn't extend to other AI companies — only to the closest direct competitors.
Read
Run the same panel