What a broken external link actually is
A broken external link is any link on your site that points off your domain to a URL that no longer works. Sometimes the target page was deleted and the server returns a 404. Sometimes the whole domain is gone and the DNS lookup fails outright. Sometimes the server is up but throwing a 500 or 503. And sometimes nothing "fails" cleanly, the request just hangs and times out, which stalls whatever is trying to load the page, human or crawler.
This is a distinct problem from broken internal links. You own both ends of an internal link, so a 404 there is almost always a config or migration mistake you can fix immediately. External links depend on someone else's site staying online and unchanged, which you cannot control. That is why external link rot needs its own recurring process instead of a one-time cleanup.
Why it matters, without overselling it
Google's own guidance (John Mueller has said this multiple times in Search Central office hours) is that a handful of broken outbound links is not a meaningful direct ranking factor. Google is not going to demote a page because it links to a dead resource. Do not panic-fix every broken link on the theory that it is costing you rankings, that is not the mechanism.
The real costs are indirect but compound over time:
- User trust and conversion. A visitor who clicks a citation or resource link and lands on a 404 forms an immediate impression: this site is not maintained. On a commerce or lead-gen page, that friction shows up in your conversion rate even if it never shows up in rankings.
- Crawl budget. Googlebot follows outbound links it discovers on your pages. Every dead external URL it requests is a wasted crawl cycle. On small sites this is negligible. On large sites with heavy outbound linking (resource hubs, citation-heavy blogs), it adds up and competes with budget you would rather spend on your own pages.
- Cumulative site-quality perception. One dead link is nothing. A site where a meaningful share of outbound links are dead reads, to both users and to Google's quality systems, as a site that is not actively maintained. Broken links are one of the easiest, most visible tells of neglect.
- Reputational exposure. If a citation link now resolves to a parked domain or spam page (a common fate for expired domains that once had inbound citations), you end up unintentionally linking to something you would never link to on purpose.
Broken external links are a hygiene and trust issue with second-order SEO effects, not a direct penalty trigger. Treat them like typos: they will not get you deindexed, but enough of them and the page reads as neglected.
How the problem actually plays out
The diagram below shows the two paths a broken external link creates: what happens when a real visitor hits it, and what happens when a crawler hits it. Same root cause, two different costs.
How to actually find them
You are not clicking through every link on your site manually, except maybe for a final spot-check. Use tools built for this:
- Screaming Frog SEO Spider. Crawl the site, then filter the "External" tab by status code and export everything that is not a 200. This is the standard first pass for sites of any size and it is free up to 500 URLs.
- Ahrefs or Semrush site audit. Both have a dedicated broken links report that separates internal from external and flags the status code for each. Useful if you already pay for one of these tools, since the check comes included.
- Google Search Console crawl stats. Settings, then Crawl Stats, shows host status and some crawl anomalies, though GSC is more useful for confirming Googlebot is not hitting a wall on your own domain than for enumerating dead external links. Treat it as a cross-check, not your primary method.
- W3C Link Checker. A free, no-install option for checking a single page's links, including external ones. Good for a quick one-off check, not built for crawling a whole site.
- Server log analysis. Pull your raw access logs and filter by the Googlebot user agent to see what the crawler is actually requesting and how those requests resolve, as opposed to what a third-party crawler simulates.
- Manual spot-checks. After an automated pass, hand-check the highest-traffic pages and citation-heavy content. Automated tools miss redirect chains that technically return a 200 but land on an unrelated page, like an expired domain bought by a parking service.
Status codes you will run into, and what to do about each
| Status | What it means | Recommended fix |
|---|---|---|
| 404 | Page not found, most common broken-link status | Check if the page moved (try the domain's search or sitemap), update the URL, or find a replacement resource |
| 410 | Gone, the site explicitly says this content was removed permanently | Do not wait for it to come back. Remove or replace immediately |
| 500 | Server error on their end, could be temporary | Retest in a few days before touching the link. If it persists past a week or two, treat like a 404 |
| 503 | Service unavailable, often maintenance or overload | Usually temporary. Recheck before editing anything |
| Timeout | No response within the request window, server may be down or overloaded | Retest from a different tool or network. If consistent, treat as dead and replace |
| DNS error | Domain does not resolve at all, often expired or abandoned | Almost always permanent. Replace or remove, do not wait |
| SSL error | Certificate expired, misconfigured, or self-signed | Site may still be live over plain HTTP. Verify manually before deciding, some are just cert neglect, not a dead site |
Fixing what you find
Work through your list in priority order, highest-traffic and highest-conversion pages first, and pick one of four fixes for each link:
- Update to the current live URL. The most common fix. Sites restructure and the content usually still exists on the same domain at a new path. Search the target domain directly rather than assuming the resource is gone.
- Replace with an archive.org Wayback Machine snapshot. If the original source is gone for good but the content was valuable enough to preserve the citation, link to a saved snapshot instead. This is a legitimate, widely used practice, and it is honest about what it is: an archived copy, not the live page.
- Remove the link and the surrounding citation. If the resource was never essential to the sentence, cut it. Do not leave orphaned phrases like "as shown here" with no link attached, rewrite so the sentence reads fine on its own.
- Find an equivalent live replacement. If the point of the link was to back a claim, swap in a current source that serves the same purpose. This is often better than an archive link because it sends readers somewhere actually maintained.
Whichever you pick, retest the link after editing. Fixing a link, publishing, and never confirming the new URL resolves is a genuinely common mistake.
What a good process looks like
This is not a one-time cleanup, it is recurring maintenance, the same category as checking for 404s on your own site. A workable cadence for most sites:
- Run a full external-link crawl quarterly. Monthly if your content leans heavily on third-party citations (research roundups, news commentary, resource lists).
- Prioritize by traffic and conversion value, not by link count. One broken link on your highest-converting landing page matters more than fifty buried in a five-year-old post nobody visits.
- Set up ongoing monitoring rather than relying purely on scheduled crawls. Some site audit platforms flag new broken links between full crawls, catching problems faster.
- Keep a simple log of what you fixed and how. It saves time on the next audit when you are trying to remember whether a flagged link was already handled.
- Prioritize high-traffic and high-conversion pages first
- Retest a link after you fix it, do not assume it worked
- Use an archive.org snapshot when the original source is gone but the citation still matters
- Run a full crawl on a recurring schedule, not just when someone complains
- Rewrite the surrounding sentence if you remove a link, do not leave a dangling reference
- Panic-fix links on the assumption they are directly hurting your rankings
- Treat a single 500 or 503 the same as a 404, retest before editing
- Link to a domain that expired and got bought by a parking or spam service without checking first
- Rely on one tool alone, crawlers and log analysis catch different things
- Ignore low-traffic pages forever, batch them into your quarterly pass instead
FAQ
Do broken external links directly hurt my rankings?
How is this different from broken internal links?
How often should I check for broken external links?
What if the target page returns a 500 or 503 instead of a 404?
Is it okay to link to an archive.org snapshot instead of the live page?
Where this fits in your broader technical SEO work
Broken external links are one input into a larger picture of site health: crawlability, internal linking, page speed, structured data, and content quality all interact. Fixing dead outbound links on your top pages is a quick win you can knock out in an afternoon with the right crawler. Building a recurring process around it is where it compounds into better site quality over time.
Our Advanced SEO Audit covers broken links, crawl budget, site architecture, and the rest of the technical checklist in one pass, with prioritized fixes instead of a raw list of problems.
Claude Vincent is a technical SEO consultant focused on crawlability, rendering, and AI-search visibility. He writes the field guides and case studies at SEO ProCheck, with a bias toward the durable, unglamorous work that decides whether search engines and AI answer engines can actually read and cite a site.
About SEO ProCheck
Technical SEO consulting and GEO strategy with 20 years of enterprise experience. Case studies, resources, and tools for search and AI visibility.
Work With Me
Technical SEO audits, GEO strategy, site migrations, and international SEO. Hourly consulting for teams who need hands-on support, not just reports.
Subscribe to our newsletter!
