# AEO Check Verify that a selected page can be accessed, understood, and quoted by answer engines and AI assistants (ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews). ## Instructions ### Before checking - Inspect the existing page, page settings, site settings, and custom code without rewriting content. - Read the Settings block first and use it as the source of truth. - Clearly distinguish between what can be verified from Framer and the live URL, and what needs manual testing in answer engines. ### Access - Check that the page is published, not password-protected, and not set to noindex. - Check robots.txt on the live domain for rules that block the crawlers in the AI crawler policy. Report each crawler as allowed, blocked, or not mentioned. - Check custom code and meta tags for directives that block AI use (e.g. noai, noimageai, nosnippet, max-snippet:0). - Check that the page appears in the sitemap. - Check that the canonical URL points to this page and not to another version. ### Content in the HTML - Check that the key content is present as text in the rendered HTML, not only inside images, videos, embeds, or canvas elements. - Check content inside tabs, accordions, carousels, and variant-based components. Content that only exists after interaction may not be read by crawlers. Mark it for manual review when this cannot be verified. - Check that important facts (price, features, requirements, compatibility) are written as text, not only shown in screenshots. ### Answerability - For each target question, check whether the page answers it clearly and where. - Check that key sections open with a direct answer within the set direct answer length, followed by detail. - Check whether headings are phrased so a matching question maps to them (e.g. "Who is [product] for?" rather than "Built different"). - Check that the page defines its main term or product in one plain sentence near the top. - Check that comparisons, steps, and specs use lists or tables rather than long paragraphs. - Check that claims, statistics, and numbers are specific and, where relevant, sourced. - Flag vague marketing phrases that give an answer engine nothing to quote. - Check for an FAQ section when the page type benefits from one, and that each answer stands alone without surrounding context. ### Entity clarity - Check that the primary entity name is spelled and styled consistently across the page, metadata, and structured data. - Check that the page says who makes the product, who it is for, and what it does, in plain words. - Check that author or founder information is present on articles, with a link to an about page. - Check that the entity description matches how the page describes the brand. - Check links to official profiles (e.g. X, LinkedIn, YouTube) from the site, and that these are referenced in Organization structured data (sameAs). ### Structured data - Check for JSON-LD in page or site custom code and list every schema type found. - Compare found types against the expected structured data for the page type. - Validate the JSON syntax. Flag invalid JSON, missing required properties, and placeholder values. - Check that structured data matches the visible content (same price, name, questions, answers, dates). - For CMS pages, check that structured data uses dynamic values and does not repeat the same static values on every item. - Flag duplicate or conflicting schema blocks from multiple sources. - Note that FAQ rich results are limited in Google Search. FAQPage markup still helps machines read the content but should not be expected to create rich results. ### Freshness and trust - Check for a visible published and last-updated date where the page type calls for it. - Check that dateModified in structured data matches the visible date. - Flag content older than the freshness threshold. - Check that pricing, feature lists, and screenshots are current. ### llms.txt - When llms.txt is expected, check whether it exists at the domain root and lists the key pages with one-line descriptions. - If Framer cannot host the file directly on the domain, mark it for manual review and note the workaround in use, if any. - Treat llms.txt as optional support, not a ranking requirement, since answer engines use it inconsistently. ### Manual verification - Suggest running each target question in the answer engines listed in the title and noting whether the site is cited, how the brand is described, and which competitor pages are cited instead. - Do not claim a page is or is not cited unless it was tested. ### Scope limits - Do not repeat the SEO check (title, meta description, heading structure, alt text) beyond where it affects answerability. - Do not write new copy or FAQ answers unless explicitly requested. Suggest which questions are missing instead. - Do not apply fixes unless explicitly requested. When adding structured data, keep it consistent with the visible content. ## Report to the User Mark each item as passed, issue found, needs manual review, or not applicable. Report in this order: access, content in HTML, answerability per target question, entity clarity, structured data, freshness, llms.txt. For each issue, explain the impact on how answer engines read or quote the page and the recommended next step. End with: - A table of target questions: answered clearly, answered partly, or not answered. - AI crawler status per crawler. - Items that need manual testing in answer engines. - The top three fixes by impact.