Research Guides
Can Search Engines Reach Page Two? A Pagination Audit Worksheet
A collection can look complete to a visitor while hiding most of its entries from ordinary link discovery. The first twelve articles load, a visitor presses “Load more,” and another twelve appear. That interaction proves the button works in that browser. It does not establish that a search crawler can discover the second group.
This audit separates three questions: can someone reach the next set, does that set have a usable address, and can a crawler discover that address through links? It is a manual website-research method, not a claim that Colquhoun.AI currently provides an automated pagination audit. Use a small sample before proposing changes across an entire site.
1. Define the collection and the expected journey
Choose one collection, such as a blog archive, product category or public resource directory. Record its starting URL, the date, the displayed item count and the navigation pattern. Is it numbered pagination, a load-more button or infinite scroll?
Write down what a visitor should be able to do. For example: “Open the archive, reach older articles, open an article, then return to the same position.” This prevents a technically neat fix from making the actual browsing experience worse.
Do not crawl every combination of filters. An archive with dates, authors, topics and sort orders can produce far more addresses than distinct useful pages. Begin with the unfiltered collection, a middle page and its final page. Use your own site or a site you have permission to inspect, and respect access restrictions.
2. Separate visible controls from discoverable links
Google's pagination guidance says its crawlers generally discover URLs through the href attribute of anchor elements. They do not click buttons and generally do not trigger functions that require user interaction to reveal more content.[1]
Inspect the next-page control. A link such as <a href="/guides?page=2">Next page</a> exposes a destination. A button whose click handler requests more records may not expose one. This is a discovery distinction, not a blanket judgment against JavaScript: Google also documents crawlable links inserted by JavaScript when they use an anchor with a valid href.[2]
For a load-more or infinite-scroll interface, ask whether there is also a paginated series of directly accessible URLs. The attractive interface can remain; the audit is looking for a reliable route to the underlying content. Record what you actually observed rather than labeling every scripted control “uncrawlable.”
3. Open the next address directly
Copy the destination into a fresh tab. It should display the intended slice without requiring you to visit page one first. Check the address after redirects, the response status and a few identifiable entries.
Then reload it. If page two becomes page one, or only works after an earlier click, the address may not represent the state it promises. Also test the browser's back button as a usability check. A visitor who loses their place may abandon a long collection even when the underlying pages are technically discoverable.
Google recommends a unique URL for each page in a paginated series and advises against using fragment identifiers, such as #page=2, for page numbers.[1] A fragment-only change is not a dependable substitute for a separate page URL in this audit.
4. Check canonical signals without collapsing the series
Page two often contains different items from page one. Do not assume that every page in the series is a duplicate just because they share a heading and layout.
Google's pagination guidance recommends each page have its own canonical URL rather than declaring the first page canonical for the entire sequence.[1] Record the canonical address on your sample pages and compare it with the content they actually contain.
A canonical tag is still a signal, not proof of Google's selected canonical or indexing outcome. If you own the site, use Search Console evidence separately to investigate those outcomes. Keep the technical observation and the search-engine observation in different columns.
5. Keep page numbers separate from filters
Pagination and filtering solve different problems. Page two extends a collection; a price sort may rearrange the same collection; a topic filter may create a smaller one. A single rule applied to all query parameters can accidentally block useful pages.
Google warns that faceted navigation can generate very large URL spaces, consuming crawl and server resources.[3] Before changing robots rules or indexing directives, classify each parameter by purpose. Decide which filtered collections deserve independent search visibility and which exist only to help visitors browse.
Do not deploy a broad parameter-blocking rule from this worksheet. Have the site's owner or developer test proposed rules against representative URLs. Blocking crawling and preventing indexing are different objectives, and a crawler must be able to retrieve a page to observe a page-level noindex directive.
6. Test the end of the collection
A collection should end somewhere. Check a clearly nonexistent page number with a single safe request, rather than generating a long sequence. Does it return a genuine not-found response, repeat the last page forever, or redirect to an unrelated landing page?
Google's faceted-navigation guidance calls for HTTP 404 responses for nonexistent pagination URLs and nonsensical or empty filter combinations.[3] Log the actual status, not merely the text “Nothing found.” A friendly error message can accompany a correct status code.
Copyable audit worksheet
| Question | Evidence to record | Follow-up if unresolved |
|---|---|---|
| Is the next set discoverable? | Control type and destination href |
Inspect rendered links |
| Does the address work independently? | Final URL, status and sample entries | Test direct load and reload |
| Is the slice distinct? | Entries on pages one and two | Review duplicate/repeated results |
| Are canonical signals consistent? | Canonical for each sampled page | Review template logic |
| Are filters treated separately? | Parameter purpose and visibility decision | Test rules before release |
| Does the sequence end? | Nonexistent-page status | Review pagination bounds |
A fictional example
A resource archive contains 60 entries, displayed twelve at a time. Its load-more button works, but page two has no exposed link. Directly opening /resources?page=2 shows entries 13–24; its canonical incorrectly points to /resources.
The useful finding is not “48 pages are missing from Google.” The audit has not measured indexing. It is: “The interface does not expose a next-page link, while a distinct second slice exists and declares the first slice canonical.” The proposed repair is to expose sequential links and review per-page canonicals, then test direct loads and navigation before release.
That precise finding is easier to verify, implement and cite than an invented visibility score. Preserve the worksheet, timestamps and sample URLs so the next reviewer can repeat the observations.
Sources
- Google Search Central: Pagination, incremental page loading, and their impact on Google Search. Checked November 12, 2026.
- Google Search Central: Link best practices for Google. Checked November 12, 2026.
- Google Crawling Infrastructure: Managing crawling of faceted navigation URLs. Checked November 12, 2026.