Find out whether AI assistants can read, understand and cite your site
AI readiness is the degree to which a site can be read, understood and cited by assistants such as ChatGPT, Claude, Perplexity and Gemini. Path Ranker measures what makes a page read and cited by AI assistants and picked as the answer: access, clear facts, direct answers and a form an agent can fill in. On paid plans, it asks the assistants themselves, shows whether your brand appears and checks which audience questions the site already answers.
How it works
Reads the site the way an assistant does
It reads robots.txt, llms.txt, the sitemap, the HTML the server delivers and the markup of the crawled pages.
Measures each page and simulates an agent
It checks 10 citability signals, 16 GEO checks and 7 AEO checks per page. An agent with the ChatGPT-User identity walks the site in a real browser up to the form, without submitting anything.
Asks the assistants (paid plans)
It writes buyer questions without your brand name and asks ChatGPT, Gemini, Claude and Perplexity, with live web search. The report shows who cited your site and who was cited instead.
What it checks
Access and discovery
AI crawlers that get into the site, llms.txt (title, summary and links) and llms-full.txt, sitemap, canonical, text in the server HTML and markup on the home page. Content that only appears with JavaScript becomes a finding.
Per-page citability
Up to 25 pages, 10 signals each: a direct answer in the first paragraph (12 to 70 words), 300 words or more, facts, lists and headings, questions, markup, author or date and a declared language.
GEO audit
Up to 15 pages, 16 checks: an answer right after each heading, heading hierarchy, quotable statements of 8 to 30 words, a direct definition, a consistent brand name, markup that matches the visible content and a date. It points out template issues, present on 70% or more of the pages.
AEO: being the answer
Seven checks per page: a question as a heading, a 20 to 70 word answer right below it, an opening with the answer and a definition, FAQPage or QAPage when there are questions, lists or numbered steps and an update date.
Facts AI needs to find
What the company does, where, for whom, how to hire it, the offer, the price, the legal identity (a tax ID such as CNPJ, NIF or EIN) and opening hours. Every fact found comes with the page and the excerpt it came from.
Agent journey
In a real browser, with the ChatGPT-User identity: discover, understand, find the offer, find the contact and fill in the form with test values, timing each step. A captcha, a custom select, a pop-up on entry and steps over 5 s become findings.
Agent protocols
It looks for /.well-known/ucp, the ai-plugin.json manifest, the A2A agent card, WebMCP, llms.txt and schema.org actions with a target (buy, reserve, schedule). A file only counts if it is valid JSON with the keys of that protocol.
AI citations and content plan
On paid plans: brand mentions and site citations in each answer, the domains cited instead and a weekly content plan for the questions with no citation. Weekly tracking repeats the same questions (up to 8) and sends an e-mail when mentions or citations fall 25 percentage points.
Audience questions
For up to 5 topics from the offer pages, it looks up what people type into the search engine (autocomplete) and the "People also ask" box, and checks which questions some page already answers, with a question heading, an FAQ or a definition. The unanswered ones become findings in the action plan, with the fix: an FAQ section with 40 to 70-word answers and FAQPage markup.
What it does not do
- No signal guarantees that an AI will cite your site. The probe records the answers of that day to those questions, not a fixed position.
- The agent never submits the form: it fills it in, checks it and stops before sending. What happens after submission is not tested.
- Citability covers up to 25 pages; the GEO audit and AEO, up to 15 crawled pages. When no browser is available, the agent journey runs on the HTML only and the report says so.
- The probe uses your own AI API keys, billed by the providers, and asks 5 questions per analysis.
- Audience questions are looked up at analysis time and tend to change over time, by country and by language. When the site is behind a login, the lookup is skipped.
Which plans include it
Starter, Pro and Enterprise get access and discovery, citability, the GEO audit, AEO, the facts, the agent journey and the agent protocols, in the full analysis or in the "AI (GEO and AEO)" group. The AI assistant probe and weekly tracking are on all three paid plans and use your own AI API keys. Audience questions are on all three paid plans too. The Free plan does not include this area.
See pricingFrequently asked questions
- GEO is getting your site ready to be read and cited by AI assistants. AEO is getting each page ready to be the answer that search engines and assistants lift out of it: the featured snippet, "People also ask", voice answers and AI Overviews. Path Ranker gives each one a score.
- No. It fills the fields with test values (such as agente-teste@example.com) to see whether an agent can, and stops before sending. Nothing reaches your inbox.
- It is one of the discovery signals: when it is missing, the report asks for a /llms.txt with a title, a summary, the important pages and the contacts. It is one signal among several; page citability and clear facts weigh more in the score.
- The probe uses your own API keys (OpenAI, Gemini, Anthropic or Perplexity), saved under Integrations and stored encrypted. Usage is billed by the providers to your account, and the Integrations screen shows the AI usage of the month.
- Only a link the answer attributes to your site. Appearing among the pages the assistant consulted is not enough; that is recorded separately.
- From the search engine itself: what it suggests when someone starts typing about your site’s topics and the questions in its "People also ask" box, looked up at analysis time. The report keeps only the questions and suggestions and marks which ones the site already answers. Your content follows what people actually ask, not just what an AI suggests.
Other features
Bing analysis
Your business on Bing Maps and your site in Bing search, which also feeds Copilot
Core Web Vitals
Real Chrome users and Lighthouse on mobile and desktop, side by side
Structured data
schema.org in JSON-LD, microdata and RDFa checked against the official vocabulary
E-E-A-T and trust
28 experience, expertise, authority and trust signals, each with the page it came from
SEO and AI goals
Ten adjustments that tend to bring clients, each with the number, the goal and its state
Monitoring and alerts
Scheduled analyses, regression alerts, downtime, and expiring certificates or domains
Keyword research
Search ideas with volume, difficulty, cost per click and intent, ready to track
Google Business Profile
A 0-100 score, local competitors, a geogrid and an alert when a profile detail changes
Google rankings
Where your site shows up in Google for the keywords that matter, and how that changes
Privacy and security
LGPD and GDPR, consent, headers, certificate, DNS and exposed credentials
Reports
PDF, PIN-protected links, your own brand and the company report, in 3 languages
AI and search bots
18 AI and search bots tested live, with what each one receives
Search Console and GA4
Google clicks, indexing and sessions in the report, including visits from AI assistants
Semantic SEO
Subjects, pillar pages, entities and passages that stand on their own
Technical SEO
Crawl, status codes, canonical, robots.txt, sitemaps, hreflang and JavaScript rendering
See your site through the eyes of an AI assistant
Run an analysis and get the least citable pages, the missing facts and the agent journey step by step, each point with its fix.
Try it free