Blog
Winning the AI answer
Where buyers ask, what gets cited, and how to tell whether AI is actually reading your site. Written for founders who are also the entire marketing department.
-
The three kinds of AI crawler, and only one means a human asked
Training, retrieval and user-triggered fetches are three different events with almost the same names. Separate them and the third class stops being an analytics stat and becomes a demand signal.
-
What to put in robots.txt for AI engines
Blocking GPTBot does not stop ChatGPT citing you. Blocking OAI-SearchBot does. A per-bot walkthrough, the file we actually serve, and five things that will bite you.
-
Why one big page cannot be cited
An engine cites a URL, not a site. Our own landing page is 1.4 MB of everything we have to say, on one URL, so it makes the point better than any example I could invent.
Miravi is an AI CMO for early-stage founders.
It keeps a brain on your product, watches which AI engines read and cite you, and drafts the pages that put you in the answer. Early access is invite-only.
Request an invite