How do you handle broken external links on sites you do not control?

Why outbound links break more often than internal ones

An internal link only breaks when your own team renames, moves, or deletes something, and you usually know when that happens. An external link depends on decisions made by strangers: a redesign that changes every URL, a domain that quietly expires, a content pruning pass, a paywall added to a formerly public page, or a company that gets acquired and folds its site into the buyer's. None of those events send you a notification. The older and more evergreen your content, the more of these you accumulate, because a post written years ago has had years for its citations to decay.

The failure modes are also more varied than a simple not found response. Some destinations return a hard 404 or a 410. Others redirect to the site's homepage, which loads fine but no longer contains what you cited. Some domains get bought and now serve a parked page full of ads, or worse, a redirect to an unrelated product. And sometimes the URL still returns 200 but the content at that address has been replaced with something entirely different. A crawler catches the status codes cleanly; a human is needed for the last case. That is why an outbound audit should report the final destination URL and page title, not only the response code. Related: When to Run a Full Link Audit

Keep reading: When to Run a Full Link Audit, How a Site Crawl Actually Works, Reading a Redirect Chain Map. See how DeadLinkr helps you on demand link and redirect audits for seo.

Running an outbound check and reading the results

Configure the crawl so it follows internal links fully but only fetches external URLs without following the links on those pages. Most audit tools call this checking external links without crawling them. Because you are hitting hundreds of unrelated hosts, per-host rate limits and generous timeouts matter more than on an internal crawl, and any 5xx or timeout should be retried before it is counted. Then group the results by destination domain. One dead domain very often explains dozens of rows, and a single decision (replace, archive, or remove every link to that host) clears them all at once. Related: A Pre Launch Link Checklist

Sort the remaining list by the value of the source page, using traffic or conversions from your analytics, so the first hour of fixes lands where readers actually are. Look separately for redirects whose final domain differs from the original domain; those are usually the domain-was-sold cases. And review 200 responses whose page title contains words like parked, for sale, or coming soon, or whose title changed drastically from what you originally linked. Those are working links that are functionally broken. Related: How Broken Links Waste Crawl Budget

Deciding what to do with each broken link

There are five reasonable outcomes for a dead outbound link: point it at an archived copy of the original page, find where the page moved on the same site, replace it with a better current source, remove the link but keep the sentence, or delete the sentence entirely. The right choice depends on what the reader expected when clicking. A citation for a study or a quote wants the archived original, because the exact text matters. A link to a tool or product wants the tool's new home or a comparable alternative, since an archived version of a download page helps nobody. Blanket-replacing everything with archive links is tempting but lazy; archived pages can be slow, partially rendered, and confusing to readers who do not know what they are looking at.

Links to parked, hijacked, or repurposed domains should be removed immediately, without waiting for the rest of the batch. Your page is now sending readers, and a link signal, to a destination you would never have chosen. For affiliate or commercial destinations, confirm that the rel attribute (sponsored or nofollow) is still appropriate after the replacement. Keep a short log of what was changed and why, so the next audit can skip items that were already decided and so a new editor can understand past choices.

Keeping outbound links from becoming a recurring chore

A quarterly outbound check is a reasonable baseline for a content site with a large archive; a site that publishes several times a week may prefer monthly. The useful comparison is not the raw count but the difference from the previous run: which failures are new, which were already known, and which were marked as intentionally left alone. An editorial policy helps upstream too. Prefer linking to stable addresses (a site's homepage, a canonical documentation page, a permanent identifier for a paper) over deep pages that get reorganized, and avoid linking to pages that already look temporary. Related: Auditing a Large Site Without Missing Pages

Prioritize by page value rather than by row count. A dead citation in a post that gets a few visits a year is not worth an editor's afternoon; it may be a signal to prune or consolidate that post entirely. Set a simple threshold that the team can actually keep: fix outbound failures on high-value pages within a week of the audit, and batch the rest into a monthly editing session. Consistency beats heroics here. A site that clears its list every quarter never faces the thousand-row cleanup that makes people give up.