Pagination: Page Series That Google Finds
Pagination is the splitting of a long list – a category, a blog, a search result – across numbered follow-on pages. It is unavoidable and, at the same time, one of the most common sources of wasted crawl budget and poorly indexed product pages. This guide shows how to build page series so every page is found without flooding the index.
1. What pagination is – and is not
A page series is an ordered sequence: /shoes, /shoes?page=2,
/shoes?page=3. Each of these is a distinct URL with distinct content. That sets pagination
apart from two neighbours:
- Faceted navigation filters and sorts the same set – it produces variants, not a continuation.
- Duplicate content exists when two URLs show the same content. Paginated pages show different items and are therefore not duplicates.
This is precisely the point most often mishandled: page 2 is not a duplicate of page 1 and must not be treated as one.
2. rel="prev"/"next" is history
For years the official advice was to chain follow-on pages with <link rel="prev"> and
rel="next". Google confirmed in 2019 that it had not used these signals for
years. Anyone still investing effort in these tags is optimising for a dead signal. They do no
harm, but they solve nothing.
Google’s stance today is plain: treat each paginated page as a standalone page reachable through an
ordinary <a href>. Everything else follows from clean fundamentals.
3. The canonical mistake that hides products
The most expensive pagination mistake: setting a canonical from every follow-on page to page 1. The logic – “page 1 is the main page” – is understandable, but the effect is damaging: Google then treats the follow-on pages as duplicates of page 1 and crawls their content less often. Products linked only from page 2 onward lose their only internal entry point.
The self-reference is correct: each paginated page canonicalises to itself.
?page=2 has a canonical to ?page=2. That keeps every page crawlable and the
products linked on it keep their internal link.
4. The four common patterns compared
| Pattern | Principle | Crawlability | Verdict |
|---|---|---|---|
| Numbered pages | Real <a href> to each follow-on page |
Full | The standard – robust and recommended |
| “Load more” button | A click loads further items | Only with a real link fallback | Fine if paginated URLs sit underneath |
| Infinite scroll | Items load automatically as you scroll | None without a URL change | Risky – see below |
| “View all” page | One page with the whole list | Full, but slow on large sets | Good with a manageable item count |
The most stable choice remains numbered pages with real links. Anything that loads items via JavaScript must have a crawlable base of ordinary URLs, or the content does not exist for Google.
5. Why infinite scroll can hide content
Infinite scroll feels modern but is the most common reason products never get indexed. The cause: Googlebot does not scroll. If content only loads through a scroll interaction that leaves no link in the rendered DOM, the bot sees only the first slice.
The fix is not to drop scrolling but to underpin it: each scroll section maps to a real URL
(?page=2) written into the address bar via the History API and reachable directly. Users
scroll, bots follow links – both get the whole content.
6. Should follow-on pages be indexed?
Opinions diverge here, and the answer depends on the page type:
| Approach | On follow-on pages | Consequence |
|---|---|---|
| All pages indexable | index, follow + self-canonical |
Recommended: products stay crawlable and findable |
| Follow-on pages noindex | noindex, follow |
Only if the pages have no search value of their own; links stay followable |
| Block follow-on pages in robots.txt | Disallow |
Wrong: also blocks following the product links |
The safe default: keep follow-on pages indexable and self-canonicalised. If you choose noindex,
make sure it is noindex, follow so the
robots directive does not cut off the product links too. Never block
follow-on pages in robots.txt – then Googlebot cannot open the page
at all to follow the links.
7. Pagination and crawl budget
On large catalogues pagination is a top consumer of crawl budget. A category with 200 follow-on pages, combined with five sort options, produces 1,000 crawlable URLs for a single category. Three levers against it:
- More items per page. 48 products per page instead of 12 cuts the number of follow-on pages to a quarter.
- Limit click depth. Page 87 sits 87 clicks deep. Additional internal linking via sub-categories pulls deep products forward.
- Canonicalise sort variants. Sorting is a facet, not new content – see faceted navigation.
8. Best practices at a glance
- Every follow-on page is reachable through a real
<a href>. - Every follow-on page canonicalises to itself, never to page 1.
- The first page ideally carries a clean URL without parameters (
/shoes, not/shoes?page=1). - Title and meta description may include the page number so snippets are not identical.
- No
rel="prev"/"next"as a substitute for real links. - Infinite scroll always has a paginated URL base.
Pro tip: Rank-O-Saur shows a page’s canonical and robots directive at a glance. Open page 2 of a category: if the canonical points to page 1, you have found exactly the mistake that keeps your deeper products out of the index.
9. Common mistakes
- Canonicalising follow-on pages to page 1. The classic – decouples deep products from crawling.
- Blocking follow-on pages in robots.txt. Prevents following the product links.
- Infinite scroll without URLs. Everything beyond the first slice stays invisible.
- First page reachable with and without a parameter.
/shoesand/shoes?page=1as two URLs create duplicates. - Identical titles on all follow-on pages. Makes it harder for Google to tell them apart; the page number belongs in the title.
- Pagination only via a JavaScript click. Without a real link in the DOM the bot never finds the follow-on pages.
10. Checklist
- Follow-on pages reachable via real
<a href>links. - Self-canonical on every follow-on page.
- First page without a
?page=1parameter. - Title/description include the page number.
noindex– if used – always withfollow.- Follow-on pages not blocked in robots.txt.
- Infinite scroll with a paginated URL base.
- Item count per page high enough to keep the page count small.
- Deep products additionally linked via sub-categories.
11. Frequently Asked Questions
Do I still need rel="prev" and rel="next"?
No. Google has not used these tags for years and confirmed this officially in 2019. They do no harm but solve nothing. What matters instead is that every follow-on page is reachable through a real link and canonicalises to itself.
Should the follow-on pages’ canonical point to page 1?
No, that is the most common and most expensive pagination mistake. Every follow-on page canonicalises to itself. If the canonical points to page 1, Google treats the follow-on pages as duplicates and crawls the products linked there less often.
Is infinite scroll bad for SEO?
Only if it has no crawlable base. Googlebot does not scroll. Without paginated URLs that are also reachable directly, the bot sees only the first slice. With a URL base, infinite scroll is fine.
Should paginated pages be set to noindex?
Usually not. Keep them indexable and self-canonicalised. Only if the follow-on pages have no
search value at all is noindex, follow defensible – the follow
is what ensures the product links are still followed.
Does page 2 of a category count as duplicate content?
No. Paginated pages show different items and are therefore not duplicates. Duplicates only arise when the same list is reachable under several URLs – for example via sort or filter parameters.
How many products belong on a page?
As many as load quickly and keep the page count small – often 24 to 48. Too few items create many follow-on pages and deep click paths; too many slow the page down and hurt Core Web Vitals.
Should I offer a “view all” page?
For manageable sets, yes – it is a good, crawlable overview. For thousands of items it becomes too large and slow; numbered pages are then the better choice.
Read next
