AEO Score
AEO (Answer Engine Optimization) checks whether your page is structured to be trusted and cited correctly - not just readable, but machine-parseable and backed by sources that hold up.
Why it exists
- Even good, readable content can get miscategorized or mistrusted by an AI system if its structured data is missing, broken, or disagrees with what's actually on the page - or if its claims aren't backed by anything checkable.
- This score looks at the layer between "readable" and "genuinely citable."
What's checked
5 checks feed this score. The last two are judged by an AI reading the page against its own structured data, not a simple diff:
| Check | Severity | What it verifies |
|---|---|---|
| Structured data | Notice | The page includes valid, correctly formatted structured data (JSON-LD) with no parse errors. |
| Schema type | Notice | That structured data declares what kind of content it is (Article, Product, FAQ, etc.) so AI can classify it. |
| FAQ schema | Notice | Q&A-style content is marked up (FAQPage/QAPage) so AI can pull question/answer pairs directly - skipped entirely on pages that aren't Q&A format. |
| Schema matches content | Warning, graded | The structured data agrees with what's actually visible on the page - mismatches get discounted as misleading, not an all-or-nothing fail. Scores full credit automatically when the page has no JSON-LD at all - that's the Structured Data check's job to catch, not this one's. |
| Sourced claims | Warning, graded | Claims are backed by named, verifiable sources - AI trusts content whose claims it can check. Only applies to Article/Recipe/Review, Dataset, and FAQ/Q&A pages - skipped entirely elsewhere (a pricing page isn't graded on sourcing). |
What moves the needle most
Same weighting every score uses:
- Critical issues count 5 points.
- Warning-level issues (unsourced claims or a schema/content mismatch) count 2.
- Notice-level polish (missing schema, no FAQ markup) counts 1.
AEO has no critical-severity checks of its own - the highest-leverage fix here is always a warning-level one: sourcing and schema consistency.
How to raise it
- Add valid JSON-LD structured data if the page has none.
- Declare the correct
@type(Article, Product, FAQPage, etc.) for what the page actually is. - Add FAQPage schema if the page answers questions in a Q&A format.
- Make sure structured data values (price, dates, author, etc.) match what's visibly on the page.
- Back claims with named, verifiable sources and attribute quotes to real people or roles.