Pagination
Three ways to split a long list into pages
Pagination means splitting content into pages. Instead of putting a product list, a list of blog posts or a long article on one screen, you split it into page 1, page 2 and so on, and let users move through them with next, previous and page number links. Google’s documentation also defines pagination as a way to browse results one page at a time through links such as next, previous and page numbers. Google (official) The same document defines two other methods as well: “load more,” where a button expands the initially shown results, and infinite scroll, where more content loads as you scroll to the bottom of the screen. Google (official)
To users, the difference between the three is a matter of preference, but to search engines it is a fork in the road. When finding pages to index, Google generally crawls the URLs in the href attribute of a tags. Crawlers don’t click buttons, and they generally don’t run JavaScript functions that change the screen only after a user action. Google (official) Yet load more and infinite scroll are usually implemented with JavaScript. Google (official) That is why people can scroll down and see page 2 products, while Googlebot stops at page 1. JavaScript SEO covers the underlying mechanics in detail.
Five rules Google wrote into its documentation
Google’s pagination document is not long. Boiled down, there are five things to follow.
- Link pages sequentially by including a link to the next page on each page with an a href tag. Google (official)
- Consider including a link back to the first page on every page in the series, so Google knows where the series starts. Google (official)
- Give each page a unique URL such as ?page=2. Google treats each of these URLs as a separate page. Google (official)
- Don’t use URL fragment identifiers (the part after #) for page numbers. Google ignores what comes after # and may not follow links that differ only after the #. Google (official)
- Don’t use the first page as the canonical page for the whole series; give each page its own canonical URL. Google (official)
The document also has one line about rel=next and rel=prev. It says Google used these tags in the past to understand the relationship between pages but no longer does, while other search engines may still use them. Google (official) Behind this one line lies an eight-year story.
rel=next/prev was retired in 2019
In September 2011, Google offered three options for pagination. Since paginated content is everywhere on the web, you could leave it as it is; if you had or planned to make a view-all page, you could follow the guidance for view-all pages; or you could add rel=next and rel=prev to each page as a hint that the pages belong to one series. Google said that with these hints it would index more accurately and show users the most relevant page, usually the first one. Search Engine Roundtable
For years afterward, these tags were considered the standard for pagination SEO. Then on March 21, 2019, Google said that after re-evaluating its indexing signals it had retired rel=prev/next. It added that, according to studies, users love single-page content, so aim for that when possible, but splitting content across multiple pages is also fine for Google Search. Search Engine Roundtable Even more surprising was John Mueller’s explanation: Google had realized it hadn’t been using the tags for indexing for years, so it removed the documentation. Search Engine Roundtable Mueller said that because Google indexes pages as it finds them, each page needs to be able to stand on its own, and that since the tags had gone unused for so long, most sites seemed to be paginating in reasonable ways regardless of these links. Search Engine Roundtable
The tags have not become useless. Bing’s Frédéric Dubut said Bing uses rel=prev/next as hints for page discovery and understanding site structure, but not for merging pages in the index or for ranking. Search Engine Roundtable In the HTML standard, next also lives on, meaning the current document is part of a series and the referenced document is the next one in that series. HTML Standard So Ahrefs recommends leaving existing rel=prev/next tags in place. There is no reason to change them, and touching them is likely to do more harm than good. Ahrefs
Canonicalizing page 2 to page 1 strands your content
A common wrong answer after the retirement news was to point the canonical tags of page 2 onward all to page 1. Ahrefs names this as the first mistake, noting that after the announcement more sites changed their implementation and ended up hurting themselves. Ahrefs Mueller had already said in February 2019, before the retirement announcement, that this was the main thing to avoid, because page 2 does not have the same content as page 1. Search Engine Journal What happens if Google accepts that canonical? As Ahrefs explains, the crawl paths to the other pages are cut, and content linked only from those pages is stranded. Ahrefs If 100 products are split across 10 pages and Google knows only page 1, then 90 of them have effectively been erased from the map.
Semrush gives the same advice. Each page in a paginated series should have a canonical tag pointing to itself, indicating that it is a unique page. Semrush If your system outputs unique URLs like /page/2/ and regular links by default, as WordPress does, the unique URL and link requirements are already met. However, WordPress core outputs canonicals only for single posts, so check the canonical for page 2 in your SEO plugin settings.
noindex is not forbidden, but it is a trade-off
Should you noindex page 2 onward? Google’s answer depends on the situation. Mueller said not to think of paginated pages as having special SEO elements, but to ask whether it would be okay if someone saw only this page on your site, and to noindex it if not. Search Engine Roundtable In 2019, in a similar discussion on Reddit, he answered to ask whether the page sparks joy and to noindex it if not, and added that if your site has only a few such series, don’t worry and let them be indexed normally. Search Engine Journal
But noindex comes at a cost. Mueller explained that if noindex stays in place for a long time, Google removes the page from the index completely and stops following its links too. Search Engine Roundtable Ahrefs sums up that adding noindex removes the page from the index so it can’t rank, and PageRank (the strength passed through links) is no longer passed on. It also warns that blocking crawling with robots.txt makes content harder to discover, strands pages and cuts off the flow of PageRank. Ahrefs Semrush advises against noindex because link equity stops flowing and discovery of new content through pagination links may be blocked. Semrush In short, even if page 2 itself isn’t worth showing in search, if some products or posts are linked only from that page, noindex is a switch that also turns off the strength of those links.
The farther from the home page, the less important Google considers it
How many clicks from the home page is page 250 of a 300-page list? Mueller explained that the home page is the most important, things linked from the home page are generally important too, and the farther a page is from the home page, the less important Google tends to consider it. Search Engine Journal So his recommendation was to reduce the number of pages: cross-link by category or tag to create several series of just a few pages each by type, and maintain a balanced hierarchy that is neither too flat nor too deep. Search Engine Journal In other words, 30 categories of 10 pages each is better than 300 pages in a single line. This is also a matter of internal link design and crawl budget.
Filter and sort pages are a different problem
Sort URLs (by price, by popularity) and filter URLs (by color, by size) are often confused with pagination. What Google’s documentation recommends handling with a noindex meta tag or robots.txt is these variant URLs, not page numbers. To keep variants of the same result list from being indexed, it advises blocking unwanted URLs with noindex or using robots.txt to discourage crawling of particular URL patterns. Google (official) If you treat page 2 and sort variants as one group and block both, you get exactly the stranding problem described above.
Common traps on Korean sites
Some Korean e-commerce platforms and bulletin board templates handle page numbers with JavaScript functions, so the links have no href or only a #. Others have only a load more button and no regular page links. As explained above, Googlebot doesn’t click buttons and ignores what comes after #, so in lists like these, products and posts beyond page 1 may never be discovered by Google. The quickest check is whether you can right-click a page number in the list and open it in a new tab, and whether the URL in the address bar changes for each page.
Long articles split across multiple pages follow the same rules. If each page has a unique URL and the pages are connected by regular links, Google indexes each page separately. As Google says, users prefer single-page content, so unless there is a real reason to split it, keeping it on one page is simpler.
References
- Pagination, incremental page loading, and their impact on Google Search – Google Search Central
- Google: Rel=prev/next Is Not An Indexing Signal Anymore; Help Docs Removed – Search Engine Roundtable
- Bing Uses rel=next and rel=prev For Discovery But Not Merging Pages – Search Engine Roundtable
- HTML Living Standard, Link types next and prev – WHATWG
- Google To Solve Pagination SEO Issue: Google’s three options in 2011 – Search Engine Roundtable
- Google Shares Guidance on Pagination for SEO – Search Engine Journal
- Google On Pagination: Would It Be Okay If The Page Was Only Accessible Via Your Site – Search Engine Roundtable
- SEOs Are Breaking Pagination After Google Changed Rel=Prev/Next, Here’s How to Get It Right – Ahrefs
Sep.2026By Studio JT