How AI engines choose which sources to cite
When someone asks ChatGPT, Gemini or Perplexity a question, the engine does not answer from memory alone. It retrieves candidate passages from the web and its index, ranks them by relevance and trust, and synthesizes a response that cites a handful of sources. This is retrieval-augmented generation, and it is the mechanism that decides whether your brand appears. Six layers determine whether you are selected.
Crawl access.
AI crawlers like GPTBot, ClaudeBot and PerplexityBot have to reach your pages. If your robots rules or a missing llms.txt block them, nothing downstream matters.
Chunk-level retrievability.
Engines retrieve passages, not whole pages. Your content has to break into self-contained chunks that state a claim and support it without the surrounding context. A 60-word answer-first paragraph is retrievable. A five-word marketing fragment is not.
Entity clarity.
Models resolve your brand to an entity in their knowledge graph. Explicit naming, schema.org markup and sameAs links to Wikidata or Wikipedia tell the model exactly who you are and what you are an authority on.
Semantic match to intent.
The passage has to answer the specific conversational query, mapped to real buyer intent, not a generic keyword. Engines match on meaning, through embeddings, not on exact strings.
Trust and authority.
E-E-A-T signals, named author and founder entities, citations and mentions from sources the models already trust raise the probability the engine selects you over a competitor.
Freshness.
Explicit publication and update signals weight recency. Stale pages lose to maintained ones on time-sensitive queries.
Our audit scores every one of these layers, shows exactly where you fail, and our agents ship the fixes.
What the GEO audit measures
Observed AI visibility.
Your real citation rate across ChatGPT, Gemini and Perplexity, measured prompt by prompt. Share of voice against competitors, and the exact pages the engines cite instead of yours.
Crawl and structured-data coverage.
Whether GPTBot, ClaudeBot and PerplexityBot can reach your content, and how completely your pages declare themselves with schema.org types: Organization, Service, Offer, LocalBusiness, FAQPage, Person.
Chunk-level retrievability.
Whether your prose breaks into self-contained, citable passages. We score answer-first density, fact density, paragraph chunking and the anti-patterns that make content unreadable to a language model.
Entity clarity and knowledge-graph grounding.
How cleanly models resolve your brand to an entity. We check explicit entity naming, knowsAbout topical declarations and sameAs grounding to Wikidata and Wikipedia.
Authority and E-E-A-T.
The trust signals that raise your odds of being cited: named author and founder entities, credibility markers, and citations and mentions from sources the models already trust.
Topical and query coverage.
The buying-intent query clusters in your category, mapped by centrality and intent and scored against what your site actually answers today. We surface the high-value gaps your competitors own.
What the agents ship
Every gap in the audit maps to an agent that produces the fix, ready to deploy.
- ✓JSON-LD schema. Valid Organization, Service, Offer, LocalBusiness and Person markup, generated for your pages.
- ✓FAQPage blocks. Structured question-and-answer sets, each answer self-contained in 40 to 80 words, with FAQPage schema.
- ✓Answer-first rewrites. Above-the-fold direct answers and chunked, fact-dense passages that engines can extract and cite.
- ✓Entity grounding. knowsAbout declarations and sameAs links that anchor your brand in the knowledge graph.
- ✓llms.txt manifest. A clean manifest of your key pages that tells AI crawlers what matters and where to find it.
AI visibility audit vs traditional SEO
SEO measures how you rank in Google's blue links. An AI visibility audit measures whether generative engines extract, trust and cite your content in their answers. A page can rank well in Google and still be invisible to ChatGPT, because the retrieval and trust signals are different. GEO adds an AI-answer layer on top of your existing SEO, it does not replace it.
From audit to fixes: agents that do the work
Most AI visibility tools stop at the diagnosis. They hand you a score and a list of problems, then leave the schema, the rewrites and the entity work to you.
Our agents take each fix and produce it: the JSON-LD, the FAQPage blocks, the answer-first content, the llms.txt, ready to deploy. You see the full audit for free. The agents that close the gap are one click away inside the platform.
AI visibility, run by people and agents
ROI & Shine pairs senior consultants with agents that execute. The agents work at machine speed and the accountability stays human. You get the judgment of people who have run growth at scale, and the throughput of automation that does not sleep.
AI visibility and GEO, answered
What is a GEO audit?
A GEO (Generative Engine Optimization) audit checks whether your website can be discovered, parsed, trusted and cited by generative AI engines such as ChatGPT, Gemini, Perplexity and Google AI Overviews. It scores both your technical and content readiness and your real citation rate in AI answers.
What is chunk-level retrievability?
AI engines retrieve passages, not whole pages, then cite the ones that answer the query on their own. Chunk-level retrievability is whether your content breaks into self-contained passages that state and support a claim without surrounding context. It is the dominant content lever in GEO.
Why does schema.org structured data matter for AI search?
Structured data declares what your page is in a machine-readable form. Types like Organization, Service, FAQPage and Person help engines resolve your entity, trust your content and extract clean answers, which raises your odds of being cited.
What is llms.txt?
llms.txt is a manifest, similar in spirit to robots.txt, that tells AI crawlers which of your pages matter and where to find them. It helps engines discover and prioritize your most citable content.
Is the AI visibility audit free?
Yes. The full diagnosis, your two-axis score and the first fixes cost nothing and need no account. You create an account only to unlock the complete plan and run the agents.
Which AI engines do you check?
ChatGPT, Gemini and Perplexity today, with more engines added over time.
Check your AI visibility now
Run the free GEO audit. See your citation rate, your readiness gaps and exactly what to fix first.
Free. No signup. Results in under a minute.