usdigitalmarketingindex.com SEO
How should search-friendly pagination be implemented for a category with thousands of products?
Updated November 15, 2026
Give every page of the listing its own URL, link the pages to one another with ordinary anchor links in the served HTML, and let each page stand as itself with its own canonical. Those three points come straight from Google’s guidance: “Give each page a unique URL”, link pages sequentially using anchor links, and “Don’t use the first page of a paginated sequence as the canonical page” (Google’s guidance on pagination and incremental page loading). The purpose in a large category is reach. Products on page 140 are found by a crawler only if there is a path of followable links to page 140, or another route to those products. Pagination is that path, and for thousands of products it has to be built so the path is short.
Page numbering and navigation
Use a query parameter or path segment that states the page number.
/boilers/ page 1 (no parameter)
/boilers/?page=2
/boilers/?page=3
...
/boilers/?page=140
Be consistent about page one. Google’s URL guidance for ecommerce sites says to “either include or exclude ?page=1 on the URL throughout your site consistently” (Google’s guidance on URL structure for ecommerce sites). Choose one form and redirect the other to it.
Sample paginated HTML links, as they should appear in the served source of page 7:
<nav aria-label="Pagination">
<a href="/boilers/">1</a>
<a href="/boilers/?page=2">2</a>
<span>...</span>
<a href="/boilers/?page=5">5</a>
<a href="/boilers/?page=6" rel="prev">Previous</a>
<span aria-current="page">7</span>
<a href="/boilers/?page=8" rel="next">Next</a>
<a href="/boilers/?page=9">9</a>
<span>...</span>
<a href="/boilers/?page=50">50</a>
<a href="/boilers/?page=100">100</a>
<a href="/boilers/?page=140">140</a>
</nav>
What matters in that sample:
- Every page link is an
aelement with anhrefholding a real URL. The guidance explains that Google generally crawls URLs found in thehrefattribute of anchor elements, and that its crawlers do not click buttons. - Links exist in the HTML the server sends, not only after a script runs.
- Besides next and previous, the block links to pages further away, so no page is many steps from the first.
- The
rel="prev"andrel="next"values on anchors are harmless and optional. Of the older link elements in the page head, the guidance says “Google no longer uses these tags”, while noting other search engines may.
What to avoid:
| Pattern | Problem |
|---|---|
| <!–INLINECODE5–> reached only through a "Load more" button with no link | No URL for a crawler to follow |
| Page state in a fragment, <!–INLINECODE6–> | The guidance says "Don't use URL fragment identifiers" for page numbers, since Google ignores fragments |
| Infinite scroll with no paginated URLs behind it | Later products never appear in fetched HTML |
| Pagination links built by <!–INLINECODE7–> handlers | Not followable |
| Only "next" and "previous" links in a 140-page series | Page 140 is 139 steps from page 1 |
Infinite scroll and “Load more” can be kept for shoppers if each batch also corresponds to a paginated URL that exists as a real page with real links.
Crawler access to deep products
Pagination is one route to deep products. In a category of thousands, it should not be the only one.
| Route | Effect |
|---|---|
| Sub-categories | Divide the listing into shorter series, each with its own pagination |
| More products per page | Fewer pages in the series |
| Skip links in the pagination block | Fewer steps to any page |
| XML sitemap of product URLs | Discovery independent of listing depth |
| Links from related products, guides, and brand pages | Additional paths |
A crawler check plan for depth:
- Crawl the site from the home page with a crawler that does not execute JavaScript.
- Record the click depth of every product URL.
- Compare the set of products found with the catalog. Any product missing from the crawl has no followable path.
- List products deeper than the limit the team sets.
- For those, add a shorter route: a sub-category, a skip link, or more items per page.
- Repeat with JavaScript rendering on, and compare. Differences show links that exist only after scripts run.
The order of products affects which ones sit deep. If the default order is stable, the same products are always on the last pages. Sub-categories are the durable fix, because they give those products a shallow listing of their own.
Page-specific metadata and canonical
| Element | Page 1 | Page n |
|---|---|---|
| URL | <!–INLINECODE8–> | <!–INLINECODE9–> |
| Canonical tag | Self | Self, <!–INLINECODE10–> |
| Title | Category title | The guidance says pages in a sequence "can use the same titles and descriptions"; adding the page number is optional |
| Indexing | Indexable | Indexable; do not add <!–INLINECODE11–> to later pages if their links are needed for discovery |
| Introductory category text | Shown | Omit or shorten, so later pages are product lists |
| Heading | Category name | Category name, with page number if desired |
The canonical instruction is the one to check first. Pointing every page’s canonical at page one tells Google the later pages are duplicates of the first. They are not: each lists different products. The guidance says to “give each page its own canonical URL”.
The guidance also suggests linking from every page back to the first page of the collection, which can hint that the first page is the better landing page.
Sorted and filtered versions of the series are a different matter. The same guidance advises keeping “URLs with filters or alternative sort orders” out of the index, so that only one series per category is indexed.
Testing later pages
| Test | Method | Pass condition |
|---|---|---|
| Served HTML contains links | Fetch page 7 without JavaScript; read the source | Anchor links to other pages present |
| A deep page loads | Request <!–INLINECODE12–> | 200, with products |
| Beyond the end | Request a page past the last | 404, not an empty 200 page |
| Canonical | Read the tag on pages 2, 50, 140 | Self-referencing |
| Page one forms | Request <!–INLINECODE13–> | Redirects to, or is canonical with, the chosen form |
| Google's view | URL Inspection on a deep page | Crawled; rendered HTML shows products and pagination links |
| Deep product discovery | Crawl comparison from the plan above | All products found |
| Server logs | Filter Googlebot requests by the page parameter | Deep pages are being requested |
| Stability | Reload a page | Same products, unless stock changed |
Two operational points. When products are added or removed, page contents shift, so the last page number changes. Make sure the page that no longer exists returns a not-found response and that links to it are regenerated. And cache paginated pages with care, so a cached page 140 does not outlive the products it lists.
Canonical tags across a paginated series, including how they interact with filtered and view-all versions, are examined in detail in the guide to how canonical tags should be configured across a paginated listing.
The rule: a unique URL for each page, real anchor links between pages in the served HTML, a self-referencing canonical on each, and more than one route to the products at the end.
Add your comment to this answer