Start with the source and the reader’s expectation
Record the page containing the broken link, its visible wording, the surrounding sentence and the exact requested URL. Open that route to confirm the problem. Distinguish a genuine missing page from a temporary server failure, a sign-in requirement or a broken section fragment on an otherwise working page.
Capture the final response and any redirects before changing anything. A crawler report provides a starting point, not the intended replacement. Read the source sentence to determine what information the visitor was promised. Preserve that expectation in the repair record so the decision does not become a search for any URL that returns success.
Find the current resource or confirm its absence
Search the authorized content inventory for the same resource and ask its owner about recent moves. Compare the candidate destination with the promised subject, language and level of detail. A broad category containing many unrelated items may not replace a specific procedure or downloadable file.
If the intended resource still exists at a corrected address, update the source link directly. If it was genuinely moved, a redirect from its old location may also be appropriate for bookmarks and external references. Keep source-link correction and old-URL handling as separate tasks: fixing one does not automatically complete the other.
Handle missing content without disguising it
When no equivalent resource exists, decide whether to remove the obsolete reference or rewrite the surrounding passage with a genuinely useful alternative. Do not keep the same promise while directing the reader somewhere else. Escalate an essential missing resource to the content owner rather than silently substituting a sales page.
Google’s migration guidance advises against redirecting many old URLs to an irrelevant destination, such as the homepage. For permanently removed material without a replacement, a proper missing-resource response can be appropriate. This article addresses the broken source link; wider retention and redirect policy should be agreed with the site owner.
Compare two hypothetical repairs
Imagine a fictional troubleshooting article linking to a specific backup restoration checklist. The checklist moved from an old downloads directory to a new help page containing the same instructions. Correcting the article’s link to that help page preserves the promised task. A separately verified redirect can serve visitors using the old address.
Now suppose the checklist was withdrawn and no equivalent instructions remain. Redirecting the old address to a generic homepage may return a successful page but does not deliver the checklist. Revise or remove the reference and document the missing resource. These examples demonstrate destination matching, not a measured recovery in traffic or ranking.
Verify the exact route after deployment
Follow the repaired link from its source page and confirm the final content, response and language. For a fragment link, check that the target section exists and is the intended one. For a download, verify the file itself rather than accepting a login screen or HTML error returned with a successful status.
Rerun the relevant link check and search for other uses of the old address within the permitted site scope. Save the original link, replacement rationale, owner decision and verification in the website-audit action list. Keep unresolved references visible in the report. A clean status code is one technical observation; a completed repair must also preserve a useful reader journey.