Question 1

What is an AI Audit?

Accepted Answer

An AI Audit examines a public website for the structural signals AI assistants (Perplexity, ChatGPT, Gemini, Google AI Overviews) need in order to cite the site as a source. The 10-check rubric covers llms.txt presence, JSON-LD schema coverage (FAQPage / Organization / Offer), canonical links, hreflang alternates, robots.txt AI-crawler posture, sitemap reachability, OpenGraph cards, Core Web Vitals. Each check carries a weight (Critical / High / Medium); the overall score is a weighted average.

Question 2

Is the audit really free?

Accepted Answer

Yes. The first-screen score and top recommendations are free. A detailed PDF (per-check evidence + remediation steps + competitor benchmark) is gated by email-capture; we use the email only for the report delivery and optional opt-in newsletter.

Question 3

How accurate is the score?

Accepted Answer

V1 of the audit runs deterministic regex + HTTP fetches; this catches roughly 80% of the signals AI assistants actually evaluate. V2 (Q3 2026) adds an LLM-graded pass that catches the remaining 20% — content quality, citation friendliness, schema correctness. Score deltas between V1 and V2 typically fall within ±5 points.

Question 4

Can I audit a competitor's site?

Accepted Answer

Yes — the audit is non-invasive (only public HTTP fetches against the visible web). Common use case: paste your top three competitors' URLs to see which one is leakiest on AI search and where to outflank them. We don't store the audited URL beyond the session unless you opt in to the PDF report.

Question 5

How does this differ from a Lighthouse audit?

Accepted Answer

Lighthouse evaluates performance + accessibility + general SEO. The AI Audit specifically grades the surface signals that AI assistants (not Google's crawler) consume to decide whether to cite the page: llms.txt, JSON-LD coverage by type, robots.txt AI-crawler posture, hreflang. Lighthouse rarely surfaces these.

Signal	Weight
llms.txt present AI assistants (Perplexity, ChatGPT, Gemini) check /llms.txt first for canonical product context. Missing = invisible to AI search.	Critical
FAQPage schema FAQPage JSON-LD lets LLMs extract Q&A pairs as discrete citation chunks. Sites without it lose every voice-search query.	Critical
Organization schema Organization + WebSite schema feed Google Knowledge Graph and AI entity recognition. Missing = the AI doesn't know what brand the page belongs to.	High
Canonical URL Canonical link prevents duplicate-content penalty across AI and Google indexing. Missing = ranking dilution.	High
hreflang for non-English hreflang alternates tell AI assistants which language version to serve. Missing = English buyers see Russian copy, and vice versa.	High
robots.txt AI-crawler posture Allow GPTBot, ClaudeBot, PerplexityBot, GoogleOther by default unless intentionally opting out. Default-blocked = invisible to AI.	High
OpenGraph + Twitter Cards Rich-link previews when shared. Required for social authority signal that LLMs lean on.	Medium
Sitemap.xml reachable Every page a search engine should index must appear here. Missing = crawl budget wasted.	Medium
Page load < 3s LCP Largest Contentful Paint within 3s. Slower = Core Web Vitals penalty and AI summarisers time out.	Medium
Structured pricing JSON-LD Offer / PriceSpecification schema on pricing pages. Missing = AI can't quote your pricing accurately, defaults to competitor numbers.	Medium

AI Audit — is your site cite-able by AI?

10 signals, weighted by AI-citation impact

How it works, what it costs, what we store

Buyer evaluation tools