What Is Pagination?
Pagination is the practice of splitting one long collection of items across a numbered sequence of separate URLs, each holding a different slice of the set.
Pagination is the practice of splitting one long collection of items across a numbered sequence of separate URLs, each holding a different slice of the set. It is not a duplicate content problem. Page two of a category listing is not an inferior copy of page one; it contains different products, and Google treats every URL in the sequence as its own page.
Why pagination decides what gets found
Pagination is where large catalogues quietly lose their inventory. A category holding 4,000 products at 24 per page runs to 167 URLs. Everything past the first one is reachable only through the sequence, so the sequence is the discovery path.
The failure is never loud. Nothing returns a 404, nothing errors, and every page renders correctly for a human clicking through. But if the deeper URLs are canonicalised away, blocked, or built as fragments, the products sitting on them are effectively unpublished. A rank tracker will not catch it, because those pages never ranked to begin with. What you see instead is a category that converts fine and a catalogue that earns almost no long-tail search traffic. Before rebuilding anything, pull the numbers and see which sections actually earn traffic, because pagination faults concentrate at the deep end of big sites.
What Google does with a paginated sequence
- It treats each page as a separate page. Google’s pagination documentation asks you to give each page a unique URL, using a parameter such as
?page=n, because URLs in a paginated sequence are treated as separate pages. - It expects each page to point at itself. The documentation is direct: do not use the first page of a paginated sequence as the canonical page, and give each page its own canonical URL. That is a different instruction from the one that governs how canonical tags work on genuine duplicates.
- It discovers deeper pages through the links. Page three is found because page two links to it. Break that chain with JavaScript that never resolves, and the crawler stops where the HTML stops.
- It ignores fragments entirely. Google ignores the text after a
#in a URL, so page numbers written as fragments create one page in Google’s view, not forty. - It indexes items where it finds them. A product listed only on page seven is associated with page seven. If page seven is not indexable, that product has no listing page behind it.
The practical consequence is that your pagination pattern sets a hard ceiling on how much of your catalogue can be found. The pattern matters more than the volume.
| Pagination pattern | What Google can reach |
|---|---|
| Unique URL per page, each self-canonical | Every item in the sequence |
| Pages 2 onward canonical to page 1 | Page one’s items, and a strong hint to ignore the rest |
| Page numbers written as URL fragments | The first page only |
| Load-more button with no URL change | Whatever renders before the button is needed |
| Paginated set plus an XML sitemap of item URLs | Items are found, but the listing structure carries no weight |
The blunt version
Two pieces of pagination advice are still sold and both are wrong. The first is rel=prev/next. Google’s own pagination documentation states that it no longer uses these tags, while noting other search engines may still read them. Plugins keep emitting them. Audit tools keep flagging them as missing. Adding them back changes nothing in Google, and any quote that lists it as a deliverable is billing you for a tag Google stopped reading.
The second is worse, because it looks tidy. Canonicalising pages 2, 3 and 4 back to page 1 is precisely what Google’s documentation tells you not to do. Run the arithmetic before you accept it. A category of 3,000 products at 30 per page is 100 URLs; canonicalise 99 of them to the first, and you have told Google that 2,970 products live on pages you have declared redundant. Your deeper inventory does not rank badly. It stops being a candidate.
The reason this keeps happening is incentive, not ignorance. Site audit tools score duplicate title tags as errors, and paginated sequences generate them by design. Canonicalising to page one clears the red rows, the report goes green, and the client sees progress. That instinct is right when two URLs genuinely compete, which is the situation behind pages fighting each other for one query. Pagination is not that situation. Consolidating it deletes reach.
Example
Say a homeware retailer lists 2,400 items in one category at 40 per page, giving 60 URLs. A developer, clearing duplicate-title warnings from an audit, adds a canonical on every page pointing at page one. Nothing visibly breaks. Customers still browse to page 12, checkout still works, and the category page holds its ranking for the head term. Six months later, product-level search traffic has drained away and nobody can name a cause, because there was no error to find. The items on pages 2 to 60 were never removed. They were simply described to Google as belonging somewhere else, and Google took the instruction at face value.
FAQ
Should page 2 canonicalise back to page 1?
No. Google’s pagination documentation says not to use the first page of a sequence as the canonical page, and to give each page its own canonical URL instead. Pointing deeper pages at page one signals that their contents are redundant, which removes the items that appear only on those pages from consideration.
Does pagination create duplicate content?
Not in any meaningful sense. The item listings differ on every page, so the pages are not duplicates. What repeats is boilerplate: the title tag, the intro copy, the filters. That produces warnings in audit tools rather than a ranking problem, and the fix is to vary the title and meta description by page number.
Is infinite scroll bad for search?
Only when it replaces URLs rather than sitting on top of them. If each batch of results has its own crawlable address, infinite scroll is a presentation choice. If the content loads without any URL changing, a crawler sees the first batch and stops, and everything below it depends on being found some other way.
Related terms
- Heading Hierarchy — the structural signal inside a page, where pagination is the structural signal between them.
- Passage Ranking — how Google surfaces one section of a long page, the alternative to splitting it.
- Canonical Tag — the directive most often misapplied to paginated sequences, with the most damage.
If a proposal offers to fix your pagination by consolidating it, it is proposing to hide most of your catalogue. Ask what happens to the items that only appear on page nine.