AI doesn't rank your brand. It resolves it. Or doesn't. Find out if you're being chosen and what's stopping the model.
No email. No credit card. No sales call. Just a look at how AI search sees you.
This report is built on Stage30™, a resolution methodology developed by Cannes-winning creative architects at Cast Iron LA (U.S. trademark application serial no. 99719523). It comes from three decades of creative leadership across web, social, and brand architecture for clients including Honda, Hilti, and Discovery Channel. See how the score is made and the documents behind it.
AI doesn't rank. It resolves. Traditional tools measure visibility, whether your brand shows up. Stage30 measures resolution, whether the model selects your brand as the answer.
Your brand is scored across the three Stage30 pillars: Authorship (the writing, a citable position in a verifiable human voice), Entity (who wrote it and the canonical graph that proves it: named authors mapped to real people, sameAs, one resolvable entity in the knowledge graph), and Code Refinement (machine-readable structure that drops the model's inference cost to near zero). The three pillars create the Resolution outcome.
This report reads one page. The Resolution Audit, our paid build, reads your brand across the entire surface of the internet, every place a model looks to resolve you. Your own site, the knowledge graph, Wikidata, Crunchbase, third-party mentions, and the author graph behind the work. It covers the surfaces where your brand has never appeared. Those absences are where the model resolves a competitor instead.
The AI Observation Log asks every engine, Perplexity, ChatGPT, Gemini, and Claude, your brand's real buyer question live, so the report shows what the models actually return, not what we guess they would. All four are included here, nothing held back.
Reading your brand signals...
This takes about 30-45 seconds.
Every finding below names the exact research document it is checked against. You can trace any number back to its source. Most audits cannot. The Stage30 Canon Source is our live index of that research: more than 60 documents on how a machine reads a page and how a brand earns trust across the open web.
For AI to repeat a fact about you, four things have to line up:
Two of these do the heavy lifting: plain words on your page, and a trusted outside source that repeats them. Below is each fact you would lead a pitch with, and where it falls short. Green means AI can get to it. Orange means it cannot.
About the code column: labeling a fact in your code helps AI read it as stated instead of guessed. It helps. It is never what gets you picked. A claim only you make, even labeled perfectly, is still just your claim. What wins is plain words on your page and a trusted outside source repeating them.
Before anything else, the model has to fetch your page, place you as one identity, and not hit a wall. These are deterministic checks, not opinions. Passing them is the floor, not the win: a readable page still has to be confirmed off-site before AI names you.
This report asks every engine: Perplexity, ChatGPT, Gemini, and Claude. "Live" means we queried each one just now. If one can't be reached, we say so and leave it out of your score.
Set the urgent stuff aside for a moment. Even on a clean site, AI weighs three simple things before it brings you up: whether your story is clear, whether the web backs you up, and whether a machine can read your site. Here is where you stand on each, in plain terms.
Each point is a fact AI can see about you but cannot use. It knows the question. It just cannot pick you, so it picks someone else.
This free report read one page and checked the rooms AI trusts. The Resolution Audit goes the rest of the way:
Human-reviewed by J and Luis, with a 30 minute strategy session. The full audit is $500 this summer. Talk to us about the full audit.
Everything above is yours, nothing held back. Drop your email and we send the PDF plus a private link to return to this exact report anytime.
Nothing here is locked. The email just saves it to your inbox and hands you a link back.
AI doesn't rank. It resolves.
This report shows you what's stopping the model.