The first site we made citable was our own.
We build AI visibility for a living. Before we sold it to anyone, we ran the exact same method on our own site and documented the before and after. This is that record. The same one every client engagement produces.
Before: invisible to machines
The site looked finished to a human. To the machines assembling AI answers, it was close to empty. There was nothing structured to read, no brief written for the crawlers, and no stated policy on who was allowed in.
| Machine layer | Before | After |
|---|---|---|
| Structured data (entities AI parses) | None. Zero JSON-LD. | 6 linked entities: Organization, ProfessionalService, WebSite, WebPage, FAQPage, Person |
| llms.txt machine brief | Missing | Published, spec-format |
| robots.txt crawler policy | None. Access undefined. | Explicit welcome for every named AI crawler |
| sitemap.xml | Missing | Published and kept fresh |
| Canonical and social metadata | None | Canonical, Open Graph, Twitter cards, favicon set |
| AI crawlers reaching the site | Untested, undocumented | 7 of 7 verified, in writing |
Had that first version shipped as it was, it would have scored about 30 out of 100 on our own scanner: crawlers could technically reach it, but there was nothing machine-readable for them to read.
The method, applied to ourselves
Hadal mapped exactly what AI crawlers could and could not see: access, structure, entities, gaps.
Entity-level schema and an llms.txt written to the audit, not a template. The machine-readable layer said what the site never quite did.
Every one of the seven crawlers checked and documented. Not "should work". Verified, with the HTTP status in writing.
The whole before and after captured in the same report format a client receives, reproducible with a single curl command.
After: 100 out of 100, verified
Today deepoceanstudio.com carries a full entity graph, a published llms.txt, a robots policy that welcomes every AI crawler by name, and per-engine verification. All seven crawlers reach and read it. The score is 100 out of 100, and it does not depend on trusting us.
This is our own site, on purpose. Engagement #0 means we do to ourselves exactly what we do for a client, and we publish it. Our live score is always checkable, and the same before and after gets produced for every engagement, in the same format. When the first client result is public, it will sit right here next to this one.
Prefer to read the raw method first? The AI Readability Reports run the same scan across 84 real businesses.