Citation Experiments
I ask ChatGPT, Perplexity, Gemini, Copilot, and Google's AI Overviews the questions your customers are asking — then I write down exactly who gets cited. This site is where I publish what I find, experiment by experiment.
Latest research
Fresh from the lab
Finding
"The GEO landscape in late 2026: why being cited beats being ranked"
The data behind generative engine optimization: market size, AI search adoption, and the collapse of the click.
2026-10-18→Finding
"Finding #001: I tried to measure AI citations as a guest. I couldn't."
Perplexity and Copilot now require sign-in before showing any answer. What the guest-access shutdown means for citation research.
2026-10-18→How it works
What I do
Ask like a buyer
Every week I ask the big AI assistants the questions real buyers actually type. Best plumber in Austin. Best CRM for a five-person team. Same wording, same order, every time — because the moment you freestyle the questions, your data is garbage.
Change one thing at a time
Then I start messing with variables. Does a quotable opening paragraph get you cited more? Do reviews move the needle? Schema? Mentions on other sites? One change, re-run, compare. Most "GEO advice" is SEO advice with a new coat of paint. I don't believe any of it until I've tested it.
Who it's for
Who this is for
If you've ever asked ChatGPT who to hire in your industry and watched it recommend your competitor — this is for you. Small businesses that want to be the answer, not the footnote.
Follow the experiments
New findings go up here and in the RSS feed. An email digest is coming — I just haven't built it yet.
Everything here is something I actually ran. AI moves fast, so I re-run old experiments when things change and update the posts.