Should you really ignore page numbers in your URLs for pagination?

What does Google really mean by ‘not relying on the page number’?

This statement by John Mueller challenges a widespread belief: many SEOs thought that Google automatically detected pagination through patterns like /page/2/, /page/3/, etc. The reality is more nuanced. The engine does not use these numbers as a direct signal to reconstruct the order of paginated pages.

Instead, Google relies more on internal links between pages, the DOM structure, and historically on the rel="next"/"prev" tags (even though they have not been officially supported since March 2019). The number in the URL can be completely arbitrary — you could have /abc/xyz/ and /def/uvw/ — as long as the navigation signals are consistent.

Why does this confusion persist among practitioners?

Many CMS and frameworks generate URLs with a numerical pattern (/page/1/, /page/2/), which creates a mental association: ‘the system must see that 2 follows 1’. This is incorrect. Google crawls and indexes each URL independently and then attempts to reconstruct the navigation relationship through links.

SEOs have long used the rel="next" and rel="prev" tags to clarify this structure. Since their official abandonment, Google has increasingly relied on analyzing internal links and patterns of duplicate or semi-duplicate content. If your pagination is not clear through linking, Google may treat each page as an isolated entity.

What are the actual signals that Google uses to understand pagination?

The engine analyzes the internal linking: presence of ‘Next Page’, ‘Previous Page’ links, clickable numbering. It also observes a common HTML structure across paginated pages (same template, same main content area). If two pages share 80% of the DOM and are linked by a ‘Next’ link, Google infers a series relationship.

It also monitors duplication patterns: if a product list partially repeats from one page to another, with simply an offset of items, it is a strong indicator of pagination. Finally, XML sitemaps can help if you include all paginated pages with consistent priorities, but this is not a direct signal of sequence.

• Clear internal linking between paginated pages (visible and crawlable Prev/Next links)
• Consistent HTML structure and repeatable across all pages in the series
• Semi-duplicate content patterns analyzed to detect series
• XML sitemaps including all paginated pages to facilitate crawling
• No canonicalization to page 1 if each page has distinct content to index

Is this statement consistent with real-world observations?

Yes and no. On e-commerce sites or large blogs, it is observed that Google sometimes ignores pages 2, 3, 4+ if the internal linking is weak or if these pages are canonicalized to page 1. But we also see cases where Google correctly crawls and indexes paginations with exotic URLs (/archives/part-alpha/, /suite/beta/), provided that the internal links are strong.

The problem is that Mueller does not provide any quantitative details: what is the exact weight of linking vs. URL pattern? How long does it take for Google to rebuild the series logic without rel="next"/"prev"? [To be verified] There is a lack of concrete data to assess the actual impact on crawl budget or indexing of deep pages.

In what cases does this rule not apply or cause problems?

On sites with endless pagination (infinite scroll), Google may never discover pages 2+ if they are only accessible via JavaScript with no HTML fallback. In this case, the URL pattern does not matter: without a crawlable link, nothing happens. Mueller speaks of ‘navigation structure’ but does not specify if this includes links loaded in JS.

Another case: paginations with combined filters (/category/page/2/?color=red&size=M). Google might interpret each combination as a distinct page, not as a logical sequence. If you want Google to understand pagination in this context, the linking must be ultra-explicit, with links to /page/3/ and /page/1/ from /page/2/. Otherwise, you risk fragmented silo indexing.

What is the gray area to watch for here?

The real question is: how does Google prioritize paginated pages for ranking? Mueller says that the number does not serve to determine the order, but we know that Google often favors page 1 in the SERPs, even if pages 2+ are indexed. Is this a freshness bias, more internal links pointing to page 1, or a hidden algorithmic logic?

[To be verified] Google does not communicate whether pages 2+ have a ranking 'handicap'. However, in practice, they rarely rank as well as page 1, even if their content is relevant. This suggests that there is an implicit signal favoring top pages, regardless of the number in the URL. Without official data, it’s an area open to interpretation.

What should you do concretely on a paginated site?

First, audit the internal linking of your paginated series. Each page should clearly point to the next and previous one, with anchors like ‘Next Page’ or ‘Previous’. If you use clickable numbering (1, 2, 3… 10), make sure it is in pure HTML, not generated solely in JS.

Next, check your canonicals. A classic mistake is to canonicalize all pages 2+ to page 1 to ‘avoid duplicate content’. However, if each pagination page contains unique products or articles, you lose their indexing. Canonical = self-referencing on each page, unless there is truly no SEO value to index.

What mistakes should you absolutely avoid?

Do not rely on the URL pattern to ‘explain’ pagination to Google. If your CMS generates /page-2/, /page-3/ but these pages are linked nowhere or blocked in robots.txt, Google will never discover them. The number is just a human label, not a robot signal.

Also, avoid noindexing pages 2+ by reflex. It was a common practice before, but if these pages contain unique content (e.g., different products), you forgo long-tail traffic. It’s better to optimize the linking and let Google index, even if it means monitoring the crawl budget on very large sites.

How can you check that your pagination is well understood by Google?

In the Search Console, go to Coverage > Indexed, and filter by URL pattern (e.g., /page/). How many pages 2+ are indexed? If you have 50 but only 5 appear, it’s a warning signal. Then check in Settings > Crawling > Crawl Statistics: are pages 2+ being crawled regularly?

Perform an internal link test with Screaming Frog: crawl your site by following only the HTML links. Are pages 2, 3, 4+ discovered naturally? If not, your linking is failing. Finally, in Google Search (site:yourdomain.com inurl:page), check how many paginated pages come up. A significant delta between existing pages and indexed pages = problem.

• Ensure that each paginated page has HTML links to Prev/Next
• Audit the canonicals: self-referencing on each page unless justified exception
• Do not noindex pages 2+ if unique content should be valued
• Control the real indexing in Search Console (Coverage)
• Monitor the crawl budget if more than 1000 paginated pages (Crawl Statistics)
• Test crawling with Screaming Frog to detect orphan pages