Free AEO checker for AI visibility.

AEO is the practice of structuring your content so AI answer engines — ChatGPT, Gemini, Perplexity, Claude — can find, understand, and quote it directly. Where SEO optimizes for blue links, AEO optimizes for being the cited answer.

SEO helps you rank in search results; AEO helps you become the answer inside AI responses. Many fundamentals overlap (crawlability, structured data), but AEO adds answer readiness — clear Q&A pairs, definitions, and citations.

It is a repeatable technical readiness score, not a promise that an AI system will cite the page. The same retrieved HTML evaluated under the same rules produces the same result, which makes before-and-after comparisons useful.

A useful AEO page might answer one specific question with a concise definition, descriptive headings, evidence-backed claims, and structured data that matches the visible content. For example, a pricing guide can lead with a direct price range, explain the variables, cite its data, and mark up the page as an Article when appropriate.

We inspect robots.txt access signals for OAI-SearchBot, GPTBot, ClaudeBot, Claude-User, Claude-SearchBot, PerplexityBot, Google-Extended, and CCBot. These controls serve different products: for example, Google-Extended does not control inclusion or ranking in Google Search.

It's a weighted blend of five dimensions: Crawlability (20%), Structured Data (25%), Extractability (25%), Answer Readiness (20%) and Visibility (10%). 90+ is excellent, 70+ good, 50+ average, 30+ poor, below 30 critical.

Human-visible ≠ AI-readable. Common killers: no structured data, missing FAQ/Q&A pairs, thin content, or a JS-rendered shell with no server HTML. The report pinpoints exactly which.

If a page is unreachable (non-200), its Crawlability score is 0 and the total is capped at 30. That red banner means fix access first — no content tweak will help until crawlers can reach the page.

No. AEO builds on the same crawlability, relevance, authority, and user-value foundations as SEO, but focuses on making answers easier for search and answer systems to extract and cite. Most teams should improve both rather than treat them as competing disciplines.

The basic check is stateless. Optional signed-in history (a later release) stores only the URL, score and timestamp so you can track trends.

Usually a few seconds. Slow target servers, redirects, and robots.txt responses can extend the check, and bounded timeouts prevent a stalled page from hanging indefinitely.

Use several signals together: crawler access and indexation, citations or mentions in answer products where they can be observed, referral visits from AI services, assisted conversions, and technical readiness over time. A readiness score alone cannot prove that a page is being cited.

Start by making the page accessible, then answer the target question directly, support important claims with evidence, use descriptive headings, and add structured data only when it matches visible content. Recheck after each change and monitor real citations and referral traffic separately.

The tool is English-first. Several answer-readiness checks currently rely on English-language patterns, so non-English pages may score lower than they should. Broader multilingual analysis is on the roadmap.

Start with the Critical cards and access failures. Make sure the page returns HTTP 200, review the crawler groups in robots.txt, add only structured data that describes visible content, and write direct answers under descriptive headings. Then rerun the audit to see which rule-level findings changed.