How to Fix Broken Outbound Links (404, 410 and 5xx Errors)
Broken outbound links point from your pages to other websites that now return a 404 (Not Found), 410 (Gone) or 5xx (server error) response. By the end of this guide you will have a confirmed list of dead external URLs, a fix for each one (replace, archive or unlink), and a repeatable way to catch new ones before an audit does.

Why it matters
Visitors who click a dead citation lose trust in the page, often within seconds. Search engines and AI answer engines judge content partly by the quality of its references, so a page that cites sources that no longer exist looks stale and less reliable. A single dead link in a footer or sidebar template repeats on every page, which can make a small problem look sitewide.
Our audit flags every page that contains at least one external link returning 404, 410 or 5xx. With fewer than 3 distinct dead URLs it reports a warning; with 3 or more it can fail.
What you need
Step-by-step fix
Collect every broken external URL and its source page
Start with the URLs listed in the audit detail. The audit shows only the first three, so run a full crawl to get the complete list.
- In Screaming Frog, go to Configuration > Spider > Crawl and tick Check Links Outside of Start Folder and External Links.
- Crawl your homepage, then use Bulk Export > Response Codes > External > Client Error (4xx) Inlinks and Server Error (5xx) Inlinks.
The export includes the Source column, which shows which page contains each dead link.
Confirm the status code by hand
Some sites block crawlers or reject HEAD requests, so re-test each URL with a GET request and a browser user agent before you change anything. Only 404, 410 and 500–599 count as broken for this check. A 403 or 429 usually means bot blocking, and redirects (301/302) that end in a 200 are fine.
Treat a 5xx as temporary until it fails again 24–48 hours later.
curl -s -o /dev/null -L -w "%{http_code} %{url_effective}\n" \
-A "Mozilla/5.0 (Windows NT 10.0; Win64; x64)" \
https://partner.example.org/2021-reportDecide on a fix for each dead link
Work through the list and pick one action per URL:
- Replace: search the target site for the moved page, or Google
site:partner.example.org "report title", and link to the new URL. - Archive: if the page is gone for good but the citation matters, link to a Wayback Machine snapshot such as
https://web.archive.org/web/2023/https://partner.example.org/2021-report. - Swap the source: link to an equivalent, current source from a different site.
- Unlink: keep the text and remove the
<a>tag, or delete the sentence if it no longer adds value.
<!-- Before --> <a href="https://partner.example.org/2021-report">2021 industry report</a> <!-- After: moved page --> <a href="https://partner.example.org/reports/2021">2021 industry report</a> <!-- After: archived copy --> <a href="https://web.archive.org/web/2023/https://partner.example.org/2021-report">2021 industry report (archived)</a>
Update global templates and menus first
Fix links that appear in shared areas before you touch individual posts: footer, header navigation, sidebar widgets, author bios, cookie banners and email signup blocks. These areas multiply the number of failing pages, so fixing them first clears the largest share of the report. com or an old http://` version of your site.
The audit treats any URL outside your exact site origin as external.
Bulk-replace links inside page and post content
For URLs that repeat across many posts, use a database or repository search-and-replace instead of editing pages one by one. On WordPress, WP-CLI lets you preview the change with --dry-run before you apply it. Always back up the database before a real run.
# Preview wp search-replace 'https://partner.example.org/2021-report' 'https://partner.example.org/reports/2021' wp_posts wp_postmeta --precise --dry-run # Apply wp search-replace 'https://partner.example.org/2021-report' 'https://partner.example.org/reports/2021' wp_posts wp_postmeta --precise
Add a recurring link check
External pages keep disappearing after you fix them, so schedule a check. Run a monthly Screaming Frog crawl, or for static or headless sites add lychee to CI. The command below treats 403 and 429 as acceptable, which mirrors what this audit flags.
lychee --accept '200..=299,403,429' --max-concurrency 8 --exclude 'linkedin\.com' 'content/**/*.md' 'src/**/*.html'
Re-crawl and re-run the audit
Purge your CDN and page cache, then re-crawl the affected source pages in Screaming Frog using Mode > List and pasting the URLs. Confirm that none of their outbound links return 404, 410 or 5xx. Then re-run the site audit.
"
Platform-specific notes
WordPress
Edit footer and sidebar links under **Appearance > Menus** and **Appearance > Widgets**, or under **Appearance > Editor > Patterns/Template Parts** for block themes. For content, use WP-CLI `wp search-replace` or the Better Search Replace plugin (Tools > Better Search Replace) with "Run as dry run" ticked first. Avoid always-on broken-link plugins that ping every URL on a schedule, because they slow shared hosting.
Shopify
Links live in several places: **Online Store > Pages**, **Online Store > Blog posts**, product and collection descriptions, **Content > Menus** for navigation, and theme sections in **Online Store > Themes > Customize** (footer and announcement bar). Shopify has no native bulk find-and-replace for content, so edit each item or use a bulk-edit app that supports rich text.
Webflow
Check static links in the Designer: select the element and open the Link settings panel. Also check CMS rich-text fields in **CMS > [Collection]** and footer Symbols/Components, which update every page at once. Publish the site after editing, because changes in the Designer don't go live until you do.
Wix
Edit links in the Editor by clicking the text or button and choosing **Link**. Blog post links are edited in **Blog > Posts**. Wix has no bulk replace, so fix headers and footers first (they are shared across pages), then work through individual posts.
Next.js / static sites
Links are usually in MDX/Markdown files or components. Find them with `grep -rn "partner.example.org/2021-report" content src`, fix them in code, and run lychee in a GitHub Actions workflow on pull requests. Redeploy so ISR or static pages are rebuilt with the new HTML.
Check that it worked
✅ How to verify
- Run curl with -L on each replaced URL and confirm a final 200 status code.
- Re-crawl the source pages in Screaming Frog List mode and confirm the External tab shows no 4xx or 5xx Inlinks for them.
- Re-run the site audit and confirm the detail reads "Sampled outbound links all resolve." with no affected pages listed.
- Open two or three fixed pages in a private browser window, click the updated links, and confirm the destination content matches the anchor text.
⚠️ Common mistakes
- Removing links that return 403 or 429. These usually mean the other site is blocking bots, the page is not broken, and this audit does not count them.
- Fixing only the post content and missing the same dead URL in a footer widget, menu or reusable block, so every page still fails.
- Pointing a dead link at the target site's homepage. The link now returns 200, but readers lose the specific source you cited.
- Forgetting to purge the page cache or CDN, so crawlers keep seeing the old HTML with the broken link.
Is this still an issue on your site? Run a free audit — ZORO checks this and 294 other SEO, AI-visibility, accessibility and security signals, with fixes for your platform.
Check my site freeFAQ
Do broken outbound links hurt SEO?
They are not a direct penalty, but they signal unmaintained content and frustrate readers. Many dead links on a page can weaken how search engines and AI systems judge its quality and freshness. Fixing them is quick and removes that risk.
Should I add nofollow to a broken link instead of fixing it?
No. Adding rel="nofollow" doesn't change the 404 or 5xx response, so the link stays broken for users and still fails the audit. Replace the link, point it to an archived copy, or remove it.
Is a redirected external link a problem?
Not for this check. A 301 or 302 that ends on a working page passes. Updating the link to the final URL is still good practice, because redirect chains can break later.
How often should I check for broken external links?
Monthly works for most sites. Check weekly if you publish news or resource roundups that cite many sources, because those links decay fastest.
How to Fix Nofollow on Internal Links (Step-by-Step)
How to Add Outbound Links to Authoritative Sources
How to Fix Generic Page Titles Like "Home" and "Untitled"
How to Fix Missing Breadcrumbs and BreadcrumbList Schema
How to Fix Accidental Noindex Tags on Key Pages