· 12 min
When AI gets your company wrong
ChatGPT, Perplexity, and AI Overviews repeat wrong facts about companies when their sources are stale or conflict. How to audit the claims, trace them, and fix them.
Insights
Notes from the lab.
Short essays on how rankings, extracted answers, and generative citations fit in one system. Written for the person who has to explain the work upstairs.
· 14 min
Rankings held, clicks fell
Stable rankings and impressions with falling clicks often point to AI Overviews, but not always. How to test it in Search Console, what else to rule out, and what to do.
· 13 min
AI search shows up in your pipeline under other names
AI search attribution links AI answers to pipeline. Most influence lands as direct, branded, or unassigned traffic, so combine GA4 channels, self-report, and CRM data.
· 16 min
AI search statistics, and what each one measured
The most-quoted AI search statistics side by side, each with its publisher, date, sample, and denominator, and why studies of the same question disagree. Last checked 2026-11-08.
· 13 min
How to choose a GEO agency without buying a screenshot
To choose a GEO or AEO agency, ask for a baseline first: a fixed query set, an engine panel, repeat runs, and mentions, citations, and accuracy scored apart.
· 13 min
One question, several searches
Query fan-out is when an AI engine turns one prompt into several search queries and cites pages that answer those. How each engine does it and how to cover it.
· 13 min
AI search cites what others say about you
Most AI citations about a brand point to sites the brand does not own. Which third-party sources engines cite, what changed with Reddit in 2026, and how to earn them.
· 11 min
How AI Overviews choose a source
AI Overviews cite pages Google's ranking systems retrieve for the query and for related sub-queries. Ranking for the exact query is no longer enough. What to build instead.
· 11 min
AI crawlers read your HTML, not your app
Most AI crawlers, including OpenAI's, Anthropic's, and Perplexity's, do not run JavaScript. What each bot sees, how to test your pages, and what to fix.
· 10 min
llms.txt is a map, not a ranking signal
llms.txt is a Markdown file at /llms.txt that points language models to a site's key pages. What the spec requires, who reads it, and how to write one.
· 12 min
Measuring AI search visibility
Run a fixed query set on a fixed engine panel, repeat each run, log answers and sources, then score mentions, citations, links, and accuracy as separate scores.
· 13 min
robots.txt for AI crawlers
AI crawlers come in three kinds, training, search, and user-triggered, and each obeys its own robots.txt token. Block one and the others still read your site.
· 10 min
SEO, AEO, and GEO, defined and compared
SEO earns a ranked link, AEO an extracted passage, GEO a citation in a generated answer. Here is how each engine selects sources, and what each one measures.
· 13 min
What schema does and does not do for AI search
Google says AI Overviews and AI Mode need no special schema. Structured data still earns rich results and clarifies entities. What to mark up, what to skip, and whether FAQ schema still matters.
· 5 min
SEO, AEO, and GEO are one system
Rankings, extracted answers, and generative citations are three scores on one query universe. Separate programs produce conflicting pages.
· 15 min
What gets cited
To get cited by ChatGPT, Perplexity, Claude, Gemini, or Copilot, a page must be retrievable by that engine and hold a specific, attributable, stable claim.
· 11 min
Share of answer
Share of answer is the share of a defined query universe where you are the cited source. How to sample it, put a confidence interval on it, and compare periods.
Bring us the domain.
We will tell you where you are absent, which page is in the way, and what has to be true before a citation is possible.