Build two inventories with different purposes
A crawl that starts at the homepage cannot reliably list every page that has no link leading to it. Begin with a separate inventory of published pages from the content system, then compare it with pages discovered by following internal links. Include the inventory date and site area so the comparison can be repeated.
Keep the original URLs alongside any normalized comparison key. Remove fragments for page-level comparison, but do not blindly merge case differences or query parameters that may select different content. Record redirects and canonical destinations as evidence rather than assuming every similar-looking address represents the same page.
Turn missing matches into candidates
Mark inventory entries that the link-following crawl did not discover. Also inspect discovered pages whose inbound-link count is zero in the exported link graph: a crawler may have received those URLs from a sitemap or a supplied list. Neither observation alone proves that no internal link exists anywhere on the site.
Check the crawl boundaries before editing. A limit on depth, an excluded directory, authentication, a failed request or unrendered navigation can hide a valid route. Revisit a suspected source page in the browser and inspect its link destination. Label the finding uncertain when coverage is incomplete rather than reporting a definitive orphan-page total.
Decide whether the page still belongs in the site
Review the candidate with its content owner. Confirm that its information is current, useful and intended for public discovery. A superseded campaign, private resource or duplicate may require a separate retention decision. Do not promote it simply because an audit export contains a zero.
For a useful current page, identify a source where readers naturally need its information. A relevant category, service explanation or related guide may work. Google recommends internal links to important pages; that does not mean every candidate needs a sitewide footer link or a particular number of incoming links.
Work through an illustrative missing guide
Imagine a fictional maintenance site with a published guide about preparing equipment for a scheduled visit. The content inventory lists the guide, but the homepage crawl never reaches it. A check shows that the maintenance overview discusses preparation without linking to the guide. The guide remains accurate and publicly useful.
Add a contextual link in that preparation paragraph after reviewing its destination. Retain a record of the source paragraph and target URL. If investigation instead reveals an existing link hidden by the crawler’s directory exclusion, correct the audit configuration and rerun it; adding another link would not explain the original discrepancy. This example describes a method, not a real customer result.
Verify the route rather than just the count
Follow the new route from the source page and confirm that the final destination serves the intended content. Inspect that the link is a normal anchor with a usable href. Repeat the link-following crawl with documented settings and confirm that it discovers the page through the source you changed.
Save the inventory entry, coverage limitation checks, chosen source, owner decision and verification in the website-audit action list. Review Arabic and English routes separately. Discovery through an internal link is a result you can demonstrate locally; it does not establish that Google has crawled, indexed or improved the ranking of the destination.