A broken link is a small defect until it sits in the path of something important. A prospective customer can reach an error instead of a service explanation, an old article can send readers to a resource that no longer exists and internal links can continue pointing at pages removed during a redesign. Regular broken-link checks help protect both the visitor journey and the internal structure search engines use to discover a website.
Understand which links need checking
Internal links point between pages and resources on your own website. External links send visitors to other sites. Both can break, but the response differs. You control your own URLs and can usually repair or redirect them directly; an external destination may need to be replaced with another useful source or removed.
Do not limit the review to ordinary text links. Navigation, image links, downloadable files and other linked resources can also fail. The most important question is whether a visitor can still reach what the page promises.
Use a crawler for a complete site review
For anything beyond a very small website, crawling software is the practical way to find broken links at scale. A crawler follows links across the site and records response information, allowing you to identify pages that point towards errors or unexpected destinations.
Export both the failing destination and the source pages that link to it. Knowing that an old URL is broken is only half the job; you need to know where the broken reference appears so it can be corrected efficiently.
Check important journeys manually as well
Automated crawling finds technical failures, but it does not always reveal whether a working link is useful. Follow key customer journeys yourself. Test navigation, service pages, calls to action, contact routes and prominent resources. A link can return a successful response while landing on irrelevant content after a poor redirect.
Manual review is particularly valuable after a redesign or content migration. Pages may technically exist but have changed meaning, leaving old anchor text pointing to a destination that no longer fulfils the promise.
Prioritise internal broken links
When an internal link points to a missing page, first decide whether the destination should still exist. If the page was removed accidentally, restoring it may be appropriate. If it has been replaced, update the source link to the current relevant URL. Where an old address still receives meaningful traffic or external links, a suitable redirect may also be required.
Avoid redirecting every deleted page to the homepage. That often fails to satisfy the visitor's original intention. Redirect only when there is a genuinely appropriate replacement; otherwise remove or rewrite the link so the journey remains honest.
Review external links with editorial judgement
External websites change constantly. A cited guide can move, a supplier can reorganise its site or an organisation can remove a resource entirely. When an external link fails, look for the intended material at a new location or find another authoritative source that supports the same purpose.
Do not replace links mechanically with the first page that happens to load. The destination needs to remain relevant to the surrounding sentence. If the reference is no longer necessary, removing the link may be better than sending readers somewhere less useful.
Investigate repeated errors at their source
If hundreds of broken links share the same pattern, individual editing may be the wrong solution. A template, menu, component or content transformation could be generating the faulty URL. Fixing the shared source prevents repeated work and reduces the chance of the error returning.
Similarly, a migration can create systematic problems when URL structures change without a clear mapping. Group errors by pattern before assigning fixes. Technical SEO becomes much more manageable when recurring causes are separated from isolated editorial mistakes.
Do not confuse every unusual response with a broken link
Crawling reports can contain redirects, access restrictions and server responses that need interpretation. A redirect is not automatically broken, although long or unnecessary chains can create maintenance problems. Some external sites may also block automated crawlers while remaining accessible to ordinary visitors.
Verify ambiguous cases in a browser before making large changes. The purpose of the audit is to improve real journeys, not to force every report row into the same status.
Build link checking into website maintenance
Broken links accumulate as content ages. Schedule periodic crawls appropriate to the size and rate of change of the site, and run checks after migrations, major navigation changes or large content removals. Assign ownership so reports lead to repairs rather than becoming another forgotten spreadsheet.
A clean link structure helps visitors move confidently through a website and helps search systems discover the pages you consider important. The strongest broken-link process combines automated discovery, manual judgement and fixes at the underlying source. That keeps maintenance focused on useful journeys rather than merely producing an error count.