DEV Community

seotto.com
seotto.com

Posted on Edited on

Pagination and infinite scroll: an SEO checklist for the component everyone builds in an afternoon

Automated technical SEO checks that catch indexing blocks and deploy regressions — seotto.com

Pagination is one of those components that gets built quickly, works perfectly for users, and quietly hides most of a site from search engines.

The pattern is common on blogs, product categories, search results, forum threads and directories. If the only way to reach an old article or a product deep in the catalogue is through page 2, 3 or 40 of a list, then how that list is paginated determines whether crawlers can find it at all.

Here's the checklist I'd run before shipping, or when auditing an existing implementation.

1. Every page in the series has its own URL

Page 2 must be reachable at a real URL, such as /blog/page/2 or /blog?page=2. If moving to page 2 only changes state in JavaScript without changing the URL, there is no page 2 as far as crawlers are concerned.

Check: click to page 3, copy the URL, open it in a private window. Do you see page 3?

2. Pagination links are real <a href> links

Search engines generally discover URLs by following links in the HTML. Buttons with click handlers, <span>s styled as links, and <a> tags without href are not reliably followed.

<!-- Crawlable -->
<a href="/blog/page/3">3</a>

<!-- Not reliably crawlable -->
<button onclick="goToPage(3)">3</button>
<a onclick="goToPage(3)">3</a>
Enter fullscreen mode Exit fullscreen mode

You can still intercept clicks with JavaScript for a smooth experience. Just keep the real href.

3. Paginated pages are in the server-rendered HTML

If the list items and pagination links only appear after client-side fetching, you're depending on rendering for discovery. Server-render or statically generate them where you can.

Check: curl page 2 and confirm both the items and the next/previous links are in the response.

4. Each page canonicalises to itself

A common mistake is pointing every paginated page's canonical tag at page 1. That tells search engines pages 2 and beyond are duplicates of page 1, which they aren't, since they list different items, and can discourage crawling of the content linked from them.

<!-- On /blog/page/3 -->
<link rel="canonical" href="https://example.com/blog/page/3">
Enter fullscreen mode Exit fullscreen mode

The exception is a deliberate "view all" page that's fast enough to be the canonical version. That's a valid choice, but make it intentionally.

5. Don't rely on rel="next" and rel="prev"

Google has said it no longer uses these as an indexing signal. They're harmless, and other consumers may use them, but they're no substitute for crawlable links between pages.

6. Infinite scroll has paginated URLs underneath

Infinite scroll is fine for users, but crawlers generally don't scroll. The established pattern is to back infinite scroll with ordinary paginated URLs:

  • Each "chunk" corresponds to a real page URL (?page=2).
  • As the user scrolls into a new chunk, update the URL with the History API.
  • Include crawlable links to the paginated URLs somewhere in the HTML, such as a hidden-until-needed pagination nav or a "next page" link at the end of each chunk.

The same applies to "Load more" buttons: make the button a link to the next page, enhanced with JavaScript.

An infinite-scroll component can hide pages from Google the afternoon you ship it. seotto.com watches for indexing and pagination-related regressions, including URLs that stopped being linked or crawlable.

7. Page 1 has one URL

/blog, /blog/page/1 and /blog?page=1 should not all exist as separate pages. Pick one (usually /blog), and redirect or canonicalise the others.

8. Out-of-range pages return 404

If there are 12 pages, /blog/page/9999 should return a 404, not an empty list with a 200 status. Otherwise you create endless soft-404s, especially if a bug or a crawler generates high page numbers.

9. Sorting and filtering don't multiply pagination

?sort=price&page=2, ?sort=newest&page=2 and ?colour=red&page=2 can generate a huge number of URL combinations. Decide which combinations you want indexed (usually the default sort only), and canonicalise or noindex the rest. Don't block them in robots.txt if you need search engines to see the canonical tags on them.

10. Titles are distinguishable

Not critical, but helpful: add the page number to titles and headings on page 2 onward, so they aren't identical across the series.

<title>Blog – Page 3 | Example</title>
Enter fullscreen mode Exit fullscreen mode

11. Deep items are reachable in fewer clicks

If your list has 200 pages and the pagination only shows "previous" and "next", an item on page 150 is 150 clicks from the start. Help discovery with:

  • Numbered links that include jumps (first, last, and a range around the current page).
  • Category, tag or date archive pages that split the list into smaller series.
  • Links between related items.
  • An XML sitemap listing the individual items, not just the list pages.

12. Ordering is stable

If items shift between pages constantly, for example a list sorted by "popular" that reorders every hour, crawlers may see different items each visit and miss some entirely. For lists you want crawled, prefer a stable default order, like newest first.

A quick test script

A rough check that the series is crawlable via plain HTML:

url="https://example.com/blog"
for i in $(seq 1 15); do
  html=$(curl -s "$url")
  next=$(echo "$html" | grep -oE 'href="[^"]*page[/=][0-9]+"' | tail -1 | cut -d'"' -f2)
  echo "$i: $url"
  [ -z "$next" ] && break
  url="https://example.com${next}"
done
Enter fullscreen mode Exit fullscreen mode

It's crude: adapt the pattern to your markup, and handle absolute URLs. But if it can't walk from page 1 to page 15 without a browser, neither can a simple crawler.

Spotting pagination problems like canonical tags pointing at page 1, soft-404 page numbers and unreachable items is part of the checks in seotto.com.

Related reading

Infinite scroll, numbered pages or "load more": which do you use, and have you checked what a crawler actually sees?

Top comments (2)

Collapse
 
elias_0549e0e2ef8cf6ad89a profile image
RAMO •

Good list. Number 9 is the one that hurts most on product categories, since sort plus filters plus page number multiplies fast.

One store-specific thing I'd add: on a product listing the pages aren't stable. Items go out of stock or get added, and a product that was on page 3 last week is on page 4 now. If deep products are only reachable through pagination, crawlers keep chasing a moving target. A product sitemap takes most of that pressure off, so pagination is mainly for users and for passing some link signals, not the only path to discovery. And on point 8, re-check it after the catalog shrinks. A category that used to have 15 pages and now has 9 should return 404 for page 12, not a quiet empty page with a 200.

Collapse
 
seotto profile image
seotto.com •

The moving-target point is a good one, product pagination is basically a list that reshuffles every time stock changes. Leaning on a product sitemap for discovery and treating pagination as mostly for people sounds right. And yes, page 12 on a category that now has 9 pages should be a 404, that one gets missed after almost every catalog cleanup.