Ecommerce Pagination SEO - Canonical, noindex, Load More or Infinite Scroll (an Audit of Polish Stores)
RETURN_TO_BLOG
AI & SEO 14 min

Ecommerce Pagination SEO - Canonical, noindex, Load More or Infinite Scroll (an Audit of Polish Stores)

Paweł Wiszniewski
Paweł Wiszniewski
SEO & GEO Specialist · AI Engineer

Ecommerce pagination splits a long category product list into consecutive pages. For SEO, one thing matters: can Google reach every product through those pages. The rule to start with: if a category has more products than the first page holds (in practice usually over 24-48), every next page needs its own URL (e.g. ?page=2), its own canonical and an `<a href>` link in the raw HTML. Do not point the canonical of pages 2+ to page 1, do not put noindex or nofollow on them, and do not rely only on a "Load more" button or infinite scroll without page URLs.

How to set up category pagination so Google finds every product: pagination, load more or infinite scroll, canonical and noindex on pages 2+, rel=next, titles, page URLs, filters, an audit of Polish stores, a script and the most common mistakes.

In this post I show what Google says about pagination today (with screenshots of the documentation), how pagination, "Load more" and infinite scroll compare, what to set on pages 2+, how page URLs and titles should look, how pagination works with filters, the results of my check of 19 Polish online stores, a script for checking it yourself and the pitfall that happens most often in stores.

Does pagination hurt SEO at all - the first decision

No. Pagination itself is neutral and Google understands it. The problem is two situations: Google cannot move on to the next pages (no links in the HTML, a button instead of a link, the page number after a # sign), or it gets a signal that the next pages do not matter (a canonical to page 1, noindex, nofollow). Then products linked only from later category pages are found by Google more slowly or not at all.

Situation in the storeIs it a problemWhat to do
The category fits on one pageNoNothing - no pagination needed
Pages 2+ have their own URL, canonical and an HTML linkNoNothing, apart from linking products from other places too
The canonical of pages 2+ points to page 1YesGive every page its own canonical
"Load more" or infinite scroll without page URLsYes, when the category has many productsAdd ?page=n URLs and links to them in the HTML
Pages 2+ with noindex or nofollowYesRemove noindex and nofollow from pagination pages
Filter and sort URLs in the indexUsually yesPaginate only the base list, handle filters separately

What Google says about pagination in 2026

Google describes pagination in a dedicated guide for online stores: "Pagination, incremental page loading, and their impact on Google Search". The key rules in short, each with a quote from the documentation.

1. Googlebot does not click buttons. Google takes URLs from the href attribute of <a> elements. A "Load more" button or infinite scroll that pulls in products with JavaScript after a user action does not lead Googlebot any further.

Google's pagination documentation: Google's crawlers don't "click" buttons and generally don't trigger JavaScript functions that require user actions - screenshot, Nov 2026
Google's pagination documentation: Google's crawlers don't "click" buttons and generally don't trigger JavaScript functions that require user actions - screenshot, Nov 2026

2. Every page has its own canonical. Google says plainly not to use the first page as the canonical page for the whole sequence.

Google documentation: don't use the first page of a paginated sequence as the canonical page, give each page its own canonical URL - screenshot, Nov 2026
Google documentation: don't use the first page of a paginated sequence as the canonical page, give each page its own canonical URL - screenshot, Nov 2026

3. Google no longer uses rel=next and rel=prev. You can keep them (other search engines may use them), but they do not replace links in the HTML.

Google documentation: Google no longer uses rel=next and rel=prev, although other search engines may - screenshot, Nov 2026
Google documentation: Google no longer uses rel=next and rel=prev, although other search engines may - screenshot, Nov 2026

4. Titles can be the same. This is an exception to the unique-title rule: pages in one sequence can share the same title and description. Adding "page 2" is not required, but it does no harm either.

Google documentation: pages in a paginated sequence can use the same titles and descriptions - screenshot, Nov 2026
Google documentation: pages in a paginated sequence can use the same titles and descriptions - screenshot, Nov 2026

5. The page number cannot sit after a #. Google ignores the URL fragment, so category#page=2 is the same page as category for Google.

Google documentation: Google ignores fragment identifiers (the text after #), so a page number after # does not create a new page - screenshot, Nov 2026
Google documentation: Google ignores fragment identifiers (the text after #), so a page number after # does not create a new page - screenshot, Nov 2026

Plus two tips from the same guide: link from every page of the sequence back to the first page, and keep filtered and re-sorted URLs out of the index (noindex or a robots.txt rule).

Pagination, "Load more" or infinite scroll - a comparison

Choosing the pattern is a decision about the customer's convenience. For SEO, each of them can work under one condition: the next batches of products have their own URLs with links in the HTML.

Pagination, Load more and infinite scroll: what Googlebot sees, which stores each pattern suits and the condition for it to work
Pagination, Load more and infinite scroll: what Googlebot sees, which stores each pattern suits and the condition for it to work
CriterionInfinite scroll"Load more"Pagination
What Googlebot sees without extrasOnly the top of the listOnly the first batchEvery page linked with <a href>
What to add for SEOPage URLs, History API, fallback links?page=n URLs and links to themNothing
Customer orientationDoes not know how many products there areSees how many are left (on the button)Sees the number of pages and where they are
Back from a product pageOften loses the positionOften loses the positionReturns to the same page
Store footerHard to reachReachableAlways reachable
Which storesInspiration, short listsFashion, interiors, decorElectronics, pharmacies, bookstores, catalogs with hundreds of products

Google has recommended for years that infinite scroll should come with a series of regular pages linked from it:

Google Search Central blog: with infinite scroll, the site or CMS should produce a paginated series of component pages - screenshot, Nov 2026
Google Search Central blog: with infinite scroll, the site or CMS should produce a paginated series of component pages - screenshot, Nov 2026

Canonical, noindex, rel=next - what to set on pages 2+

On pages 2+ use a simple set: own canonical, no noindex, a link to the next page and to the first one. This is how it looks on page 2 of the coffee category in the bee.pl store:

Source code of page 2 of a category in the bee.pl store: its own canonical (page-2) and robots index,follow - screenshot of the page source, Nov 2026
Source code of page 2 of a category in the bee.pl store: its own canonical (page-2) and robots index,follow - screenshot of the page source, Nov 2026
Setting on page 2+What happensVerdict
Own canonical (?page=2 points to ?page=2)Google treats the page as separate and follows its links furtherYes
Canonical to page 1A signal that page 2 duplicates page 1 - Google may skip it together with its product linksNo
No canonicalGoogle picks the canonical itself, with URL parameters this can be randomBetter add your own
noindexThe page drops out of the index, and with it the path to some productsNo
nofollow (meta robots)Google does not follow the links on this page, product links includedNo
rel=next and rel=prevGoogle does not use them, other search engines mayOptional
A "View all" page as canonicalWorks only when the whole list loads fastRarely in stores
The correct pagination setup: every page has its own URL and canonical, a href links to the next and first page, and four mistakes that make products on later pages disappear
The correct pagination setup: every page has its own URL and canonical, a href links to the next and first page, and four mistakes that make products on later pages disappear

Page URLs and titles

The URL can carry the number in a parameter (?page=2, ?p=2) or in the path (/page/2, /page-2). Both work for Google, what matters is that it is stable and has no #. In my audit, 14 of 19 stores use a parameter and 5 use the path.

Three rules that are easy to forget:

  • Page 1 without a number. The "1" link in pagination should lead to the category URL, not to ?page=1. If ?page=1 exists, give it a canonical to the category URL or a 301 redirect.
  • A non-existent page is a 404. ?page=999 should not return an empty list with a 200 code or redirect to page 1.
  • A suffix in the title is optional. 16 of 19 checked stores add "page 2" to the title. It helps customers in browser tabs, but Google does not require it.

Pagination with filters and sorting

Paginate only the base category list. Once a customer picks a filter or a sort order, URLs like ?color=black&page=2 or ?sort=price&page=3 multiply very fast and eat crawl budget. For Google, the fix is the same as for filters alone: only filters with demand are indexed as separate pages, and the rest of the combinations get noindex or a robots.txt rule. I describe how to decide which filters to index in the post on faceted navigation in online stores.

Watch the order: if you block every URL with the page parameter in robots.txt and pagination uses that parameter, you also block regular category pages 2+. The rule should cover filter and sort combinations, not the page number alone.

How many products per category page

Google gives no single number. A practical rule: as many as keep the page fast on a phone (check LCP in Core Web Vitals) while the key products of the category stay at most two clicks from the home page. More products per page means fewer pages for Googlebot to walk through, but a heavier page for the customer.

Category sizeProducts per pageWhat else to do
Up to 50 products24-48, often a single pagePlain pagination or none
50-500 products24-60Pagination with HTML links, subcategories for groups with demand
Over 500 products36-60Subcategories and filters with demand as separate pages, a product sitemap

In large categories pagination alone is not enough for Google to find new products at the end of the list quickly. Links from subcategories, "new in" sections and the sitemap help - I cover them in the post on ecommerce indexing.

What my audit of 19 Polish stores showed

On 21 November 2026 I checked pages 1 and 2 of one category in Polish online stores (electronics, drugstores, pharmacies, bookstores, fashion, furniture, pet supplies). I fetched the raw HTML, the version without JavaScript, and compared it with the version rendered in Chromium. Pagination could be checked in 19 stores. In a few others, bot protection blocked access or I did not find a category with a second page, so they are not in the results.

Audit results for 19 Polish stores: link to page 2 in the raw HTML, own canonical, title, rel=next, canonical to page 1, nofollow and noindex on page 2
Audit results for 19 Polish stores: link to page 2 in the raw HTML, own canonical, title, rel=next, canonical to page 1, nofollow and noindex on page 2
Element on category page 2StoresComment
<a href> link to page 2 in the raw HTML of page 118 of 19One store shows links only after JavaScript
Own canonical15 of 19The model setup
Canonical to page 13 of 19In one case the canonical is added only by JavaScript
No canonical1 of 19Google picks the canonical itself
noindex0 of 19No store removes pages 2+ from the index
nofollow in meta robots1 of 19Google does not follow product links from this page
rel=next or rel=prev11 of 19Google does not use them, but they do no harm
A different title than page 116 of 19A "strona 2" ("page 2") suffix
Page number in a parameter14 of 19?p=2, ?page=2, ?strona=2, ?counter=1, ?start=61
Page number in the path5 of 19/strona/2, /page-2, /2/
Page number after #0 of 19-

Selected stores from the sample:

StorePage 2 URLPage 2 canonicalLink in raw HTMLNotes
bee.pl/page-2ownyesrel=next and prev
doz.pl/strona/2ownyesrel=next
zooplus.pl?p=2ownyesrel=next
morele.net/2/ownyestitle with "strona 2"
taniaksiazka.pl/page-2ownyesrel=next
eZebra.pl?counter=1ownyeszero-based numbering in the parameter
drogerienatura.pl?p=2ownyesrobots INDEX,NOFOLLOW on page 2
rossmann.pl?Page=2page 1yessame title as page 1
empik.com?start=61page 1yesrel=next and prev
swiatksiazki.pl?p=2page 1 (after JavaScript)noraw HTML without the product list
smyk.com?p=2noneyesrel=next

Key takeaways from the audit:

  • The basics are fine. 18 of 19 stores have a link to page 2 in the raw HTML, and none uses noindex or a page number after #.
  • A canonical to page 1 is still a common pattern. 3 of 19 stores have it. Large stores usually have their products in sitemaps and other categories too, so the effect can be smaller, but in a smaller store where a product is linked only from its category it can mean products Google never reaches.
  • rel=next survives in more than half of the stores. It does no harm, but it does not replace links and should not be relied on.
  • The riskiest case is a list that exists only after JavaScript. In one store the raw HTML of the category page has neither products nor links to further pages, and the canonical appears only after rendering. Google can render it, but later, and AI crawlers do not run JavaScript - more in my post on JavaScript SEO.

I check pagination, filters and canonicals across all listing templates as part of my ecommerce SEO service.

How much work - two scales

ScopeSmall store (up to 1k products, a hosted platform)Large store (over 20k products, a custom template)
Pagination audit1-2 hours (the script below on a few categories)1-2 days (crawler, server logs, all listing templates)
Canonical and robots fixA theme or SEO plugin setting, 1-2 hoursTemplate change and tests, 2-5 developer days
Replacing "Load more" with a version with page URLs1-2 days (often built into the theme)1-2 weeks (URLs, History API, device tests)
Checking the effectSearch Console after 2-4 weeksServer logs and Search Console after 2-6 weeks

Script: check pagination on your category

The script fetches pages 1 and 2 of a category without JavaScript, the way a crawler sees them before rendering, and shows the canonical, meta robots, rel=next, the title and whether page 1 has an <a href> link to page 2. It needs only Python 3.

check_pagination.py
# Usage: python3 check_pagination.py https://store.com/category https://store.com/category?page=2import re, sys, urllib.requestfrom urllib.parse import urljoinUA = "Mozilla/5.0 (compatible; pagination-check/1.0)"def fetch(url):    req = urllib.request.Request(url, headers={"User-Agent": UA})    with urllib.request.urlopen(req, timeout=20) as r:        return r.status, r.geturl(), r.read().decode("utf-8", "ignore")def meta(html, base):    head = html.split("</head>")[0]    can = re.search(r'<link[^>]+rel=["\']canonical["\'][^>]*href=["\']([^"\']+)', head, re.I)    rob = re.search(r'<meta[^>]+name=["\']robots["\'][^>]*content=["\']([^"\']+)', head, re.I)    title = re.search(r"<title[^>]*>(.*?)</title>", head, re.I | re.S)    return {"canonical": urljoin(base, can.group(1)) if can else None,            "robots": rob.group(1) if rob else "(none)",            "rel_next": bool(re.search(r'<link[^>]+rel=["\']next', head, re.I)),            "title": title.group(1).strip() if title else None}def links(html, base):    return {urljoin(base, h).split("#")[0] for h in re.findall(r'<a[^>]+href=["\']([^"\']+)', html, re.I)}page1, page2 = sys.argv[1], sys.argv[2]s1, u1, h1 = fetch(page1)s2, u2, h2 = fetch(page2)m1, m2 = meta(h1, u1), meta(h2, u2)print("Page 1:", s1, m1)print("Page 2:", s2, m2)print("<a href> link to page 2 in the HTML of page 1:", page2 in links(h1, u1))if m2["canonical"] and m2["canonical"].rstrip("/") == u1.rstrip("/"):    print("WARNING: the page 2 canonical points to page 1")if "noindex" in m2["robots"].lower() or "nofollow" in m2["robots"].lower():    print("WARNING: page 2 has noindex or nofollow")if "#" in page2:    print("WARNING: page number after # - Google ignores the URL fragment")

Note: some large stores block requests that do not come from a browser (a 403 code or a "Just a moment" page). Then check the same by hand: open page 2, choose "View page source" and search for canonical and robots, and in Search Console use URL Inspection, which shows the version Googlebot sees.

Production pitfall: a canonical to page 1 after a theme or plugin change

The most common scenario I see in audits: the store had correct pagination, then changed its theme or SEO plugin, or deployed a new listing template, and pages 2+ started pointing their canonical to page 1. Nobody notices, because the page looks the same, and the effects arrive after a few weeks: Search Console shows more URLs as "Alternate page with proper canonical tag", and new products at the end of the list get indexed more and more slowly.

How to protect yourself:

  1. 1.After every theme, SEO plugin or category template change, run the script from this post on 3-5 categories.
  2. 2.Add this test to your deployment checklist (or to automated tests, if the store has them).
  3. 3.Once a month, check in Search Console whether the number of pagination pages with "Alternate page with proper canonical tag" has suddenly grown.

The second variant of the same pitfall: a plugin adds noindex to pages 2+ "for tidiness". Nothing happens in the short term, but over time Google visits these pages less often, and products linked only from them lose their path to the index.

What not to do

  • Do not point the canonical of pages 2+ to page 1.
  • Do not add noindex or nofollow to pagination pages.
  • Do not put the page number after a # (category#page=2).
  • Do not rely only on a button or infinite scroll without page URLs and HTML links for further products.
  • Do not block the page parameter in robots.txt if pagination uses it.
  • Do not count on rel=next as the only way to connect pages - Google does not use it.
  • Do not return an empty list with a 200 code for non-existent pages (?page=999).

FAQ

Summary

Ecommerce pagination is one of those things that keep working for years once set up, as long as nobody breaks them with a theme change. A check takes a few minutes: an <a href> link to page 2 in the raw HTML, an own canonical, no noindex or nofollow, a page number without a #. I describe the order of all store work in the ecommerce SEO strategy. If you want me to check pagination, filters and indexing in your store and prepare template fixes, see my ecommerce SEO service.

/// AUTHOR
Paweł Wiszniewski – AI & Web Engineer

Paweł Wiszniewski

SEO & GEO Specialist & AI Engineer

SEO/GEO specialist (10 years) and AI engineer (3 years). I build search visibility, AI systems and automations that reduce costs and improve operational efficiency.

Signal received?


Book a free consultation

Initiate protocol. Establish connection. Let's build something loud.

> WAITING_FOR_INPUT...

Form data is controlled by E-Marketing Paweł Wiszniewski (Suwałki, Poland) and used only to reply and prepare an offer. Details and your rights: Privacy policy.