Broken Links in Published Documents: Why They Matter More Than You Think
Broken links are the fastest way for a document to look abandoned. A reader who clicks a link and gets a 404 error immediately forms an opinion about the quality of everything else in the document.
Why links break
Links break for predictable reasons, and understanding them helps prevent the problem.
**URL restructuring.** The target organisation changes its website structure. A page that lived at /resources/policy.pdf is now at /documents/policies/policy.pdf. The old URL returns a 404.
**Content removal.** The target page is deleted, archived, or moved behind authentication. This is especially common with regulatory guidance, which gets revised and the old versions are removed.
**Domain changes.** The target organisation changes its domain. All links to the old domain break simultaneously.
**Temporary resources made permanent.** A document links to a staging environment, a shared drive folder, or a temporary hosting location that was always going to be taken down.
**Link rot over time.** Studies of academic papers find that a significant percentage of links are broken within a few years of publication. Documents that are not actively maintained suffer the same fate.
The consequences by document type
In technical documentation, a broken link to an API reference, a library, or a configuration guide means the reader cannot complete the task the documentation describes. The documentation fails at its primary purpose.
In compliance filings, a link to a regulatory standard that returns 404 may cause a reviewer to question whether the standard exists, whether the correct standard was cited, or whether the filing is current.
In contracts and legal documents, hyperlinks to incorporated policies or standards are increasingly common. If the linked policy changes or the link breaks, there is ambiguity about what version of the policy was incorporated at the time of signing.
In reports distributed to external audiences, broken links damage credibility and create a poor user experience that reflects on the organisation.
Distinguishing link problems
Not all link problems are the same.
A 404 error means the resource does not exist at that URL. A 403 means the resource exists but is not accessible without authentication. A redirect chain means the URL resolves but through multiple hops, which may eventually break. A link to a page that still resolves but whose content has changed is a more subtle problem — the link works, but it no longer points to what it claimed to.
Automated link checking handles the first three cases reliably. The last case — content drift — requires human judgment.
Detection and maintenance
Automated link checking extracts all hyperlinks from a document, attempts to resolve each one, and reports on the result. For documents that are updated periodically, this check should be part of the update cycle.
For static documents intended to remain accurate over time, the better practice is to avoid linking to external resources that may change, and instead to quote or attach the relevant content directly. Where linking is necessary, prefer stable permanent URLs — DOIs for academic papers, archived versions for web pages, and versioned URLs for regulatory documents.
What to do about existing broken links
For each broken link, the options are: find the correct current URL and update it, remove the link if the content is no longer relevant, or replace the link with direct quotation of the essential content.
Do not leave broken links in place simply because fixing them requires effort. A document with known broken links and no remediation plan is a document whose author has accepted that it does not work.