I test what AI cites so you don’t have to guess.

I’m EG. I help brands get cited, recommended, and trusted by AI engines. I publish AEO/GEO research, live experiments, audits, and practical playbooks for improving visibility in ChatGPT, Perplexity, Gemini, and other answer engines.

Most advice in this space is theory, written before anyone looked at the data. My approach is the opposite: I ask the engines real questions, record every source they cite, and publish the raw results — including the parts that contradict the conventional wisdom.

Why hands-on?

AI answer engines are a moving target. What gets cited shifts by engine, by model tier, and by whether the model searched the live web or answered from memory. You can’t reason your way to the truth from first principles — you have to measure it. So I do, and I show my work: the queries, the engines, the exact sources, and what changed.

What you’ll find here

• Original citation findings — real queries against real engines, every cited source recorded.
• Patterns, not platitudes — what cited pages actually have in common, plain enough to act on.
• Playbooks — the structural moves that get a page cited, and your brand seen.

This site practices what it preaches

Every article is built to be cited: a direct answer up top, question-shaped headings, extractable tables and lists, an FAQ, and clean structured data. If the methods work, this site should show up in the answers — a test I’m happy to run in public.