Part of the Technical SEO Audit series. For the full picture, start with Technical SEO audit: the complete guide and checklist.
Broken links pile up quietly. A product gets discontinued, a blog post is renamed, a partner site shuts down, and the links pointing at them stay in templates, menus and old articles. No one notices until a visitor clicks one.
This guide covers what counts as a broken link, which ones matter most, how to find all of them in one crawl, and how to fix them so they stay fixed.
What is a broken link?
A broken link is a link that points to a URL that doesn’t work: it returns a 4xx client error (usually 404 Not Found or 410 Gone), a 5xx server error, or no response at all because of a timeout, DNS failure or SSL error.
They come in three kinds, and each one needs a different fix:
| Kind | Example | Who fixes it |
|---|---|---|
| Internal link to a broken page | Blog post links to /pricing-2023/, which now returns 404 |
You: update the link or restore the page |
| External link that broke | Article cites a study whose page was removed | You: update or remove the link |
| Broken image | Product photo URL returns 404 after a CDN change | You: fix the URL or restore the file |
Do broken links hurt SEO?
They waste more than they damage. Google’s Page indexing report documentation says plainly that “404 responses are not necessarily a problem, if the page has been removed without any replacement”, and its guide to HTTP status codes confirms that 4xx codes other than 429 “have no effect on crawl rate.”
The real costs are elsewhere:
- Lost pages. Google says it doesn’t index URLs that return a 4xx status, and removes already-indexed URLs that start returning one. If a page that ranked now returns 404 by mistake, its traffic is gone.
- Server errors are worse. 5xx responses “prompt Google’s crawlers to temporarily slow down with crawling”, and URLs that persistently return a server error are removed from the index.
- Wasted links. An internal link to a dead page sends visitors and crawlers to a dead end instead of to a page you want them to reach.
- Trust. A help article full of dead links looks abandoned, to readers and to anyone deciding whether to link to you.
Which broken links should you fix first?
Not all broken links are equal. Work in this order:
- Pages that still get traffic and now return errors. These lose visits every day. If you connect Search Console or Google Analytics, you can see exactly which ones.
- Internal links to 4xx and 5xx pages, especially in navigation, footers and templates, where one broken link repeats on every page.
- Pages that don’t respond at all. Timeouts and connection errors often mean a server or firewall problem, not a missing page.
- Broken images.
- Broken external links, starting with the pages that get the most traffic.
How do you find broken links on a whole site?
Checking pages one at a time in a browser extension shows you one page. To see every broken link, crawl the site: follow every internal link, record the status code of every URL, and keep track of which page linked to which.
Crawlens does that in one crawl and groups the results into these checks:
| Check | What it flags | Severity |
|---|---|---|
| Internal pages return 4xx | Crawled URLs that return 404, 410 or another client error | Critical |
| Internal pages return 5xx | Crawled URLs that return a server error | Critical |
| Pages did not respond | Requests that timed out or failed (DNS, connection, SSL) | Critical |
| Pages link to broken internal URLs | Pages with links to any of the above, with the list of broken targets | Critical |
| Broken images | Pages with internal images that fail to load | Warning |
| Broken external links | Pages with links to external URLs that return an error or don’t respond | Warning |
| Pages with search traffic return errors | URLs that earned clicks from Google and now return an error | Critical |
| Landing pages with visits return errors | URLs where visitors from any channel land, now returning an error | Critical |
Two settings decide what gets checked after the crawl. Check images is on by default and requests every internal image. Check external links is off by default, because it’s slower on link-heavy sites; turn it on in the crawl settings when you want external links checked. It requests up to 10,000 external URLs per crawl, trying a lightweight HEAD request first and falling back to a normal GET for servers that reject it.
The last two checks need Search Console or Google Analytics connected. They’re the ones to read first, because they put a number on what each broken URL is costing.
How do you fix broken links?
The fix depends on why the target broke.
| Situation | Fix |
|---|---|
| Page moved to a new URL | 301-redirect the old URL to the new one, then update internal links to point straight at the new URL |
| Page removed, close replacement exists | 301-redirect to the replacement, then update the links |
| Page removed, no replacement | Leave it returning 404 or 410, and remove or change the links to it |
| Page deleted by mistake | Restore it |
| 5xx or no response | Check the server logs for that URL; it’s an application, hosting or firewall problem |
| External page gone | Link to the new location if it moved, an archived copy or another source, or remove the link |
Two habits keep this list short:
- Fix the link, not just the target. A redirect makes the old URL work, but every internal link that still points at it now goes through a redirect. Crawlens flags those separately as internal links that point to redirects. See redirect chains for why that matters.
- Fix templates once. If the same broken link appears on hundreds of pages, it’s in the header, footer, sidebar or a shared component. Open any affected URL, check its Outlinks tab, and fix the component instead of the pages.
To find where a broken URL is linked from, open it and look at its Inlinks tab, which lists every page linking to it with the anchor text.
How do you stop links from breaking again?
- Redirect when you change URLs. Renames, category restructures and CMS migrations are the biggest source of broken links. Map old URLs to new ones before launch; the site migration checklist covers this step by step.
- Re-crawl after every change that touches URLs, and use Compare to confirm the broken links are gone without new ones appearing.
- Schedule crawls so new breakage shows up within days, not when a customer reports it.
Checklist
- Full crawl with Check images on, and Check external links on
- Search Console or Analytics connected to find broken URLs that still get traffic
- Broken URLs with traffic restored or redirected first
- Broken links in templates (header, footer, sidebar) fixed in the component
- Internal links updated to point straight at the final URL
- Removed pages without a replacement left as 404 or 410, with links removed
- 5xx and no-response URLs checked in the server logs
- Broken external links updated or removed
- Re-crawl and compare to confirm
Frequently asked questions
Do broken links hurt SEO?
A 404 on its own doesn't harm the rest of the site: Google says 404s aren't necessarily a problem when a page was removed with no replacement. The cost is in what the broken link wastes: visitors hit a dead end, crawlers spend requests on it, and links into the page stop passing value.
What is the difference between a broken link and a 404 page?
A 404 page is a URL that returns a 404 status. A broken link is a link on one of your pages that points to a URL like that, or to one returning a 5xx error or no response. You fix a broken link by editing the page that contains it.
Should I redirect every 404 to the homepage?
No. Redirect a URL only when the content moved or a close replacement exists. A page removed with no replacement can keep returning 404 or 410, and the links pointing to it should be updated or removed.
How do I find which pages link to a broken URL?
Crawl the site with a tool that stores the link graph. In Crawlens, the issue for broken internal links lists each linking page with its broken targets, and the Inlinks tab of any URL shows every page that links to it.
How often should I check for broken links?
After every site change that touches URLs, such as a redesign, migration, CMS change or content cleanup, and on a regular schedule otherwise. Monthly works for most sites; large sites that publish daily benefit from weekly crawls.