Which items should a paginated or infinite-scroll list mark up for carousels beta?

Definition: For Google's structured data carousels beta, a paginated category page should mark up the entities listed on that page in an ItemList. An infinite-scroll page should focus its markup on the entities initially loaded in the viewport.

Key facts

• Google says to mark up all items on a summary or category page, then add an ItemList to each subsequent page of a paginated category with that page's entities. Google's carousel documentation gives this page-by-page rule.
• For infinite scroll, Google's beta guidance says to focus on the entities initially loaded in the viewport. The same documentation uses the initial viewport as the scope.
• The marked-up page is a summary or category page containing information about at least three entities and links to their separate detail pages. Google's beta requirements set both conditions.
• Google's ItemList definition requires at least three itemListElement.item elements, with a 1-based position for each item. The beta property definitions specify these requirements.

What belongs in a paginated page's ItemList?

Each paginated category page gets an ItemList containing the entities that the page lists. As of 11 September 2026, Google's beta documentation applies the rule to the initial summary page and to each later page in the sequence. The summary or category page must contain information about at least three entities and link to other pages with more information about those entities.

The url inside each item points to the canonical standalone detail page. Google requires all list URLs to point to different pages on the same domain, and its guidance excludes anchor links inside the summary or category page. A page can mix supported entity types when its scenario calls for it, but every URL still identifies a separate detail page.

How should pagination be exposed to Google?

Google defines pagination as links such as “next”, “previous”, and page numbers that lead to pages displaying one page of results at a time. Its pagination documentation, last updated 10 December 2025, recommends linking each page to the following page with an <a href> element and giving each page a unique URL. Google generally crawls URLs in the href attribute of anchor elements, while its crawlers generally do not activate buttons or user-action JavaScript that updates page contents.

The page URL and the ItemList content should therefore describe the same page of results. Google says that fragment identifiers are ignored for pagination URLs, so a page number should be represented by a distinct URL rather than only by text after #.

What belongs in an infinite-scroll page's markup?

For infinite scroll, the beta guidance narrows the working set to entities initially loaded in the viewport. Mark up those entities in the page's ItemList; later content that appears only after scrolling is outside the specific focus stated by this guidance. The page still needs the beta's summary-page structure, including links to standalone detail pages for the listed entities.

Google's general pagination guidance describes infinite scroll as loading more content when a user reaches the end of the page. It also says that “load more” and infinite scroll are generally implemented with JavaScript and that crawlers generally do not trigger JavaScript functions requiring user actions. This makes the initially available HTML and structured data especially important for the items selected for the markup.

What does the markup look like?

ItemList is the container for the list, and each entry is a ListItem with an item object and a position. Schema.org's ItemList definition says that ListItem is used for ordered lists when additional context or position matters, and that markup order alone is insufficient to indicate order.

{
  "@context": "https://schema.org",
  "@type": "ItemList",
  "itemListElement": [
    {
      "@type": "ListItem",
      "position": 1,
      "item": {
        "@type": "Restaurant",
        "name": "Trattoria Luigi",
        "image": ["https://example.com/photos/1x1/photo.jpg"],
        "url": "https://www.example.com/trattoria-luigi"
      }
    }
  ]
}

Google's example uses the same ItemList, ListItem, position, item, name, image, and URL shape. The complete list must contain at least three item objects for this beta, and the items represented should be the entities on the relevant page or initially loaded viewport.

Related terms

• Pagination
• Infinite scroll
• Structured data

FAQ

Should a later paginated page repeat the items from the summary page?

No. Google's beta documentation says each subsequent page gets an ItemList containing the entities listed on that page.

Should infinite scroll markup include every item a user might load?

Google's beta guidance focuses on the entities initially loaded in the viewport. Its separate pagination documentation says crawlers generally do not trigger user-action JavaScript that updates page contents.

Does the ItemList position start over on every paginated page?

Google documents position as a 1-based number for the item's position in the carousel. The beta page does not specify whether later-page positions restart or continue across the sequence.

Sources

1. Google Search Central, "Structured data carousels (beta)", 2026-09-11
2. Google Search Central, "Pagination, incremental page loading, and their impact on Google Search", 2025-12-10
3. Schema.org, "ItemList", 2026-09-16

---

Check how a page reads to answer engines. Open the GEO score analyzer