Direct answer
Infinite interaction should be backed by stable paginated URLs and crawlable links so people and crawlers can reach items without executing a custom gesture.
Give every batch a URL that works directly
Infinite scroll can remain the visitor experience while pagination provides discovery. A crawler should be able to open a later batch without first scrolling or pressing a button. Google generally discovers links through anchors and does not perform the user actions needed to trigger load-more controls.
Give each page a distinct URL, link sequentially with href anchors and use a self-canonical for each page of the sequence. Do not canonicalize every batch to the first page or use fragments as the only batch identifier.
<link rel="canonical" href="https://example.com/products/?page=2">
<nav aria-label="Product pages">
<a href="/products/">Previous page</a>
<a href="/products/?page=3">Next page</a>
</nav>Define states before adding the scroll interaction
| State | Expected behavior | Manual check |
|---|---|---|
| Open page 2 directly | Show batch 2 and its product links | Paste its URL into a new session. |
| Append the next batch | Preserve the existing items and useful navigation | Check focus, duplicates and the Back button. |
| Request fails | Keep already loaded content and expose a retry | Simulate a failed batch request. |
| Final batch | Stop requesting more and state the end | Confirm the footer stays reachable. |
| No such page | Use the chosen truthful empty or missing state | Test a page number beyond the sequence. |
Worked example: a 240-item catalog
Suppose a collection displays 24 items per batch. The acceptance inventory contains ten batch URLs and 240 product destinations. For a stable test fixture, request each batch directly and compare the union of product IDs with the fixture inventory. This catches an offset bug that repeats batch 2 or silently drops the final items.
On a changing catalog, use stable ordering and document how insertions or removals affect batch boundaries. Keep sort orders and filter combinations separate from the base discovery test. A successful test with yesterday's inventory is not proof that today's inventory is complete.
Verification checklist for infinite scroll SEO
- Request the first, middle and final batch without any preceding navigation.
- Check that the actual HTML or rendered output contains the expected item links.
- Verify the self-canonical and real href to the following batch.
- Test scrolling, keyboard operation, reload and back/forward navigation on mobile.
- Inspect a representative deep product in Search Console and compare it with server evidence.
- Record coverage gaps separately from confirmed indexing problems.
Common mistakes
- Hiding all later items behind scroll events.
- Canonicalizing every page to page one.
- Creating unbounded calendar paths.
Limitations
- Collection churn affects reproducibility.
- Crawler depth is not a fixed click-count rule.
Primary sources
- Pagination and incremental page loading — Google Search Central; Documentation reviewed 2026-09-26; 2026-09-26.
- Make your links crawlable — Google Search Central; current 2026-08-19; 2026-08-19.
- Understand the JavaScript SEO basics — Google Search Central; 2026-03-04; 2026-08-14.
- Web Content Accessibility Guidelines (WCAG) 2.2 — W3C; 2024-12-12; 2026-08-14.