How to Find Orphan Pages and Decide What to Do With Each One

An orphan page is a page on your site that no other page links to. It may sit in your sitemap, collect a trickle of visits from an old newsletter or even still rank for something, but neither visitors browsing your site nor crawlers following links can reach it. Google's link guidance is blunt about it: "Every page you care about should have a link from at least one other page on your site."

The difficulty is that orphans are invisible by definition. A crawler that follows links never finds a page that nothing links to, so a crawl report on its own cannot show them. You need a second list of URLs that comes from somewhere other than your links, and a careful comparison.

This guide shows how to build those lists, compare them, confirm each candidate and decide what it deserves: a link, a merge, a redirect or retirement. It closes with how to link pages back in so they stay found, and how to stop new orphans appearing.

What counts as an orphan page

Strictly, an orphan page has no internal links at all. In practice what matters is whether a crawler can follow a link to it. Google's link best practices say it can generally only crawl a link that is an <a> element with an href attribute, and its pagination guidance adds that Google's crawlers don't "click" buttons. So several kinds of page are effectively orphaned even when a person can reach them in a browser.

The last row is not an orphan, but it is worth reviewing in the same pass. Treating navigation links and contextual links separately is Backlnk's editorial distinction, not a Google rule.

Why internal links matter for discovery and context

Google's guide to how Search works describes URL discovery in three ways: pages Google already visited, pages found when Google extracts a link from a known page (for example a category page linking to a new post), and URLs submitted in a sitemap. Links also carry meaning. Google says it "uses links as a signal when determining the relevancy of pages and to find new pages to crawl", and that anchor text tells people and Google what the linked page is about.

Google's sitemap documentation connects the two: if your pages are properly linked, meaning every important page can be reached through some form of navigation, Google can usually discover most of your site. It also notes that on large sites it is harder to make sure every page is linked from at least one other page. A sitemap helps search engines discover URLs, but it "doesn't guarantee that all the items in your sitemap will be crawled and indexed".

Common causes of orphan pages

Orphans are rarely created on purpose. They are a side effect of normal site work:

• Navigation changes. A redesign or menu cleanup drops the only link to a deep page.
• "Load more" and infinite scroll. Archives that replace numbered pages with a button leave older items without crawlable links. Google recommends giving each page in a sequence its own URL and linking pages sequentially.
• Links that aren't links. Cards, tabs or menus built as buttons or with onclick instead of <a href>.
• Migrations. Pages move to new URLs, old internal links are not updated, and the old URLs return 404 instead of redirecting.
• Campaign pages. Landing pages built for ads or e-mail that were never linked from the site.
• Automatic sitemaps. A CMS, shop or documentation system adds every published page to the sitemap but not to any menu, category or hub.
• Removed categories or tags. The listing page disappears; the items it listed stay.

How to find orphan pages: compare three lists

The method is a comparison. One list shows what your links can reach; the others show what exists or gets traffic. Anything in the second group but not in the first has no crawlable link that the crawl could find.

1. Crawl the site from the homepage. Use a crawler that follows <a href> links on your own host and records every URL with its status. If menus or links are inserted by JavaScript, crawl the rendered page. Let the crawl finish: URLs a crawl never reached because it hit a page limit are unknown, not orphaned.

2. Export every sitemap URL. Start from the Sitemap: lines in robots.txt and follow sitemap index files to every child sitemap. (Our guide to XML sitemap best practices covers which URLs belong there.)

3. Export URLs that get impressions or visits. In Search Console's Performance report, switch the table to the Pages dimension, widen the date range (the default is three months) and export. Add landing pages from your analytics and, if you have them, URLs requested in your server logs.

4. Normalize the lists. Make protocol, host (www or not), trailing slashes, letter case and parameters consistent with how your site writes URLs, and drop fragments. Otherwise the same page looks like two different URLs.

5. Compare. URLs in the sitemap or traffic lists but not in the crawl are orphan candidates. URLs in the crawl but missing from the sitemap are sitemap gaps worth fixing too.

6. Verify each candidate. Open it. Does it return HTTP 200, and should it be indexable? Check whether it is reachable only through scripts, nofollow links or pages your robots.txt blocks. In Search Console's Links report, open Internal links, then Top linked pages, and filter by the URL to see which pages Google knows link to it.

Sheet "Crawl", column A:      every URL the crawl reached
Sheet "Candidates", column A: sitemap and traffic URLs, de-duplicated
Formula in Candidates!B2, filled down:

=IF(ISNA(XLOOKUP(A2, Crawl!A:A, Crawl!A:A)), "Orphan candidate", "Linked")

What a crawl alone cannot show: our test

On October 31, 2026 we ran Backlnk's Internal Link Analyzer on https://www.w3.org/WAI/ with a 40-page budget on the source HTML. It discovered 1,344 internal URLs, crawled 40 of them, and found no broken internal links among the 734 links it checked. It also listed five pages linked only from navigation and one internal link marked nofollow.

What it could not do was look for orphans. robots.txt declared no sitemap and /sitemap.xml answered HTTP 404, so every URL the tool knew about had been found through a link, and the report marked the orphan check as "not assessed". That is the method's limit, not a verdict on the site: without a list from outside the link graph, there is nothing to compare.

Decide what to do with each orphan page

Not every orphan should be rescued. Decide page by page, based on whether it serves a purpose people search for and whether another page already serves it better.

Retiring is a normal outcome. Google's documentation on HTTP status codes says it treats 404 and 410 the same way: URLs that return them are removed from the index if they were indexed, and crawling of them gradually decreases.

Link orphan pages back in: anchor text and placement

Once a page should stay, give it links that both crawlers and people can use. Google's link guidance describes good anchor text as descriptive, reasonably concise and relevant to both the page it sits on and the page it points to. Its practical test: read the anchor text out of context and check that it still makes sense. The words around a link matter too, and several links chained next to each other make it harder to tell them apart.

There is no ideal number of links per page; Google says so explicitly, adding that if it feels like too many, it probably is. Our interpretation for prioritizing: pages that matter commercially should be a few clicks from the homepage and linked from at least one page that already receives traffic, so both visitors and crawlers reach them quickly.

Check your site with Backlnk's Internal Link Analyzer

The Internal Link Analyzer crawls your site from a start URL (same host, up to 40 pages per run, respecting robots.txt) and reads your sitemap from the robots.txt Sitemap: lines or /sitemap.xml. Sitemap URLs are added to the link graph and checked after the link crawl, which is how it finds orphan candidates. It reports:

• URLs listed in the sitemap with no incoming links observed in the crawled set: your orphan candidates.
• Pages linked only from navigation, kept apart from links in the main content.
• Broken links (404 or 410), links to redirecting URLs and temporary failures, each listed separately.
• Internal links marked nofollow and pages more than three clicks from the start page (a disclosed Backlnk heuristic).
• Link opportunities that quote the sentence on the source page where a link would fit, with a suggested anchor.

Its wording is deliberately careful. A bounded crawl cannot prove that a page has no links anywhere on a large site, so the report says "no incoming links observed in the crawled set" and shows exactly how many pages were crawled, skipped or blocked. Without a sitemap it marks orphans as not assessed, as in our test above, and in source mode it does not see links inserted by JavaScript (choose rendered mode for those). For sites larger than one run, analyze sections separately and use the three-list comparison for the full picture. The free full audit combines this check with Backlnk's other tools in one report sent by e-mail.

Next step: stop new orphans appearing

Run the three-list comparison once, fix what you find, then make linking part of publishing:

• Every new page gets at least one contextual link from a related, already-linked page when it goes live.
• Before removing a menu item, category or tag, check which pages only it linked to.
• Archives and listings use paginated URLs with ordinary links, not only "Load more".
• After a migration, internal links point straight to the final URLs.
• Repeat the crawl, sitemap and traffic comparison after every large site change.