By Rahul Saini, Author at Search Counsel Co. Last updated [Month] 2026. Part of our ecommerce SEO guide.

1) The modern truth about pagination

The advice you may remember from a few years ago is out of date. Two things have changed, and they change everything.

First, Google treats each paginated page as its own standalone page. Page two of a category isn’t a fragment of page one; it’s a separate URL competing for crawl budget, indexing, and rankings on its own. Second, rel=”next” and rel=”prev” are no longer used for indexing. Google retired them as an indexing signal back in 2019. You can keep them for accessibility, but they do nothing for your rankings, and relying on them is a mistake. What matters now is plain, crawlable links and correct canonicals.

Put simply: pagination isn’t bad for SEO. Poor pagination is. The whole job is making sure each useful page state is reachable, stable, and correctly signaled.

2) The pagination rules that work

For an ecommerce category split across multiple pages, this is the setup that holds up in 2026:

• A unique, stable URL per page. Use /category?page=2 or /category/page/2/, each returning a 200 status with real server-rendered HTML containing that page’s products.
• A self-referencing canonical on each page. Page two points its canonical at page two, not at page one. This is the opposite of what a lot of old guides say, and it matters (section four explains why). The canonical tag in general is covered in duplicate content and canonicalization.
• Unique title tags. “Running Shoes, Page 2” distinguishes the page in search and avoids looking like a duplicate. The H1 can stay consistent (“Running Shoes”) with a small page indicator.
• Real, crawlable link anchors. The previous and next links must be actual HTML anchors in the rendered page, not buttons that only work with JavaScript. Hide them visually if your design needs, but keep them in the HTML.
• A sensible item count. Google’s ecommerce guidance points to roughly 24 to 48 items per page. Fewer items means more pages and more crawl demand; more items means slower loading. Find the balance for your catalog.

3) Numbered vs load-more vs infinite scroll

Three patterns dominate, and they’re not equal for SEO.

If you want load-more or infinite scroll for the experience, that’s fine, but build it as a progressive enhancement on top of real pagination. Three technical pieces make it safe:

1. Update the URL as the user scrolls using the History API (pushState), so each visible section has a shareable, bookmarkable URL like ?page=3.
2. Server-render each paginated URL. When a crawler requests ?page=3 directly, the server should return full HTML with that page’s products, not an empty shell waiting for JavaScript.
3. Keep real previous and next anchors in the HTML, so Googlebot can follow the sequence without executing any scroll or click.

Skip this and you get the classic disaster: a beautiful, fast store where the crawler sees the first twenty products and nothing else. The rendering and JavaScript side is covered further in our JavaScript SEO guide.

4) The number-one pagination mistake

Here it is, the single error that quietly de-indexes catalogs: canonicalizing pages two and beyond to page one.

It feels tidy. Page two looks like a near-duplicate of page one, so pointing its canonical at page one seems reasonable. But those pages show different products. When you canonical page two to page one, you’re telling Google the products on page two don’t have their own home, so Google can drop them from the index entirely. On a large catalog, that can bury a huge share of your products. Each paginated page should self-reference instead, so every product remains discoverable. Reserve canonicalizing-to-another-page for filtered, sorted, and parameter URLs, which is a different scenario entirely.

5) Pagination and crawl budget

Pagination multiplies your URL count fast. A store with 200 categories each running ten pages deep has 2,000 category URLs before you count a single product. For a site under about 10,000 pages, this rarely causes trouble. For large catalogs, deep pagination can eat a meaningful share of crawl budget, and Googlebot starts deprioritizing thin, repetitive deep pages.

Watch for the warning signs in Search Console: lots of pages stuck in “Discovered, currently not indexed,” deep paginated pages missing from the index, or high crawl activity with a low ratio of indexed pages. If a category runs fifteen pages deep and pages eight through fifteen get no clicks, it’s worth asking whether they need to be indexed at all, though on most stores the safer fix is stronger internal linking and more items per page rather than aggressive noindexing. The broader mechanics are in our crawl budget guide.

6) How pagination fits your architecture

Pagination is one piece of a bigger question: can a crawler get from your homepage to every product efficiently? A few structural principles keep the whole store reachable.

• Keep it shallow. Important pages should sit within about three clicks of the homepage. The deeper a page, the less often it’s crawled and the less authority it receives.
• Use a clear hierarchy. Home leads to category, category to subcategory, subcategory to product, each level passing authority down. How you design that taxonomy is covered in category page SEO and subcategory and attribute pages.
• Add breadcrumbs. Breadcrumb navigation with BreadcrumbList schema reinforces the hierarchy for shoppers and search engines.
• Maintain an XML sitemap of your canonical, indexable URLs, so nothing important depends on crawlers following every on-page link, which matters most on JavaScript-heavy sites.
• Server-render your catalog. If your store is built on a JavaScript framework, make sure category and product HTML is server-rendered so crawlers see products without executing scripts.

Good pagination inside a flat, well-linked architecture is what lets a large catalog get fully crawled and indexed. General site-architecture and URL-structure fundamentals live in our technical SEO hub.

7) Pagination and AI search

The crawlability problem is even sharper for AI. Generative engines parse links, URLs, and HTML rather than simulating a user, and they don’t scroll any more than Googlebot does. A catalog hidden behind infinite scroll with no crawlable fallback can vanish from AI shopping answers entirely, even while the site looks fast and modern. The same paginated, server-rendered structure that helps Google also keeps your products retrievable by AI. We cover that side in our AI search optimization guide.

8) Common mistakes

• Canonicalizing pages 2+ to page 1. The most damaging pagination error. Self-reference each page instead.
• Infinite scroll with no fallback. Hides everything past the first batch from crawlers. Layer it over real paginated URLs.
• Product links behind buttons or scripts. If the links aren’t crawlable anchors in the HTML, Google can’t follow them.
• Relying on rel=next/prev. It’s not an indexing signal anymore. Use crawlable links and canonicals.
• Burying products too deep. Pages many clicks from home get crawled rarely. Flatten the structure and strengthen internal links.
• Inconsistent pagination links. Links pointing variously at ?page=1, the clean URL, and parameter versions confuse crawlers. Standardize.

9) Sources used

FAQ: pagination and site architecture

Where to go next

Pagination and architecture come down to one test: can a crawler reach every product through stable, crawlable HTML links? Use numbered pagination with self-referencing canonicals, put paginated fallbacks under any scroll or load-more experience, keep your structure shallow and well-linked, and server-render your catalog.

That completes the technical set. Control the filters that create most duplicates in faceted navigation SEO, handle the canonical tag itself in duplicate content and canonicalization, or step back to the full ecommerce SEO guide.