ItemList schema is structured data that tells search engines and AI assistants that a page contains an ordered or unordered list of items — the "10 best CRMs," a ranking, a roundup, a set of steps. When paired with the right page types, it can power carousel rich results in Google and, more importantly, it hands AI answer engines a clean, ranked set of entities to pull from when they assemble "best of" recommendations. If you publish listicles or comparison roundups, ItemList is how you turn prose into a machine-readable list.

What ItemList does and where it applies

An ItemList contains ListItem entries, each with a position and either a URL or a nested item. There are two common patterns. A summary list (all-in-one-page) is a ranking or roundup living on a single page, where each ListItem points to a section or an external target. A document list links each ListItem to a separate URL — useful when every entry has its own detailed page. Carousel rich results are supported when ItemList is combined with an eligible content type such as Article, Recipe, Movie, or Course; for a roundup of your own guides, each item can reference an Article or BlogPosting entity.

The position property is the part people underuse. Order carries meaning — a ranked list says "we consider #1 better than #5," and that ranking is information an AI assistant can repeat. Get the positions right and consistent with the visible order on the page.

Why ItemList is quietly powerful for AI recommendations

When someone asks an assistant "what are the best project management tools?", the model assembles an answer by pulling ranked, compared entities from sources it trusts. A page marked up as an ItemList of tools — each with a name, position, and link — is dramatically easier to lift than the same information trapped in flowing paragraphs. You're doing the structuring work the model would otherwise have to guess at. This is the same logic behind winning comparison and best-of content: give the machine a table or list it can read directly, not adjectives it has to interpret.

ItemList also reinforces internal structure. A roundup that links to your own detailed guides, each marked up in turn, builds a small connected cluster that signals topical depth — the kind of authority that gets a source cited repeatedly rather than once.

Building an ItemList page that works

Structure the page and the markup together so they mirror each other:

If your list items are entities you've described elsewhere, reuse their stable identifiers so the assistant connects the roundup entry to the full entity. Consistent identifiers through the @id property keep "the CRM ranked #3 here" and "the CRM described in detail there" recognizable as the same thing.

Common ItemList mistakes

The usual failures are structural. Positions that don't match the visible order, or that skip numbers, confuse parsers and undercut the ranking signal. Marking up a list of items that isn't actually shown on the page violates the content-visibility rule and risks a manual action. Using ItemList for navigation menus or unrelated link groups dilutes the meaning — reserve it for genuine lists of comparable items. And expecting a carousel rich result without pairing ItemList with an eligible content type leads to disappointment; validate the combination before assuming it will render.

Done right, ItemList turns your best roundups and rankings into exactly the kind of structured, quotable content that AI assistants reach for when they recommend. It's low effort for content you've already written — you're just labeling the order the machine should read.

Want to see whether AI crawlers can read your lists and parse your ItemList markup? Run the CheckMy.site scanner to check how your roundups and rankings appear to the assistants deciding which sources to recommend.