Broken internal links: dead ends on your own site

Why they happen on careful sites

• Renamed handles and slugs. The link was correct for months; the product handle changed and every link pointing at the old slug died at once.
• Unpublished or deleted pages. A collection retired or a blog post removed. The links pointing at it stay until someone notices the 404s.
• Typo-ed URLs in rich text. Hand-typed links inside product descriptions and blog posts never got tested, and a missing character is invisible until clicked.
• Staged domains. Content copied from staging with hardcoded staging-domain links that 404 in production.

What they cost

Visitors hit a dead end and trust drops. Crawlers hit the same wall and log the URL as a problem. Enough dead links and the crawl spends its budget on your failures instead of your products. Internally, every dead link is authority the page was supposed to pass that goes nowhere. And the worst ones are invisible to you: the page you are reading renders fine, the broken link is one click deep, and only a request reveals it.

Find them, then fix at the source

1. Fix the link, not just the target. Updating the target URL leaves the broken link in place to fail again; edit the page that contains the link.
2. If the page moved, redirect the old URL, but fix the links anyway, because redirects are a hop, and hops cost (see redirect chains).
3. Sweep templates and rich text after the fix: a renamed handle usually broke the same link in several places, and the check reports the first page it appeared on, not all of them.
4. Re-run and watch. Links break when content changes, which is exactly when nobody is looking. The daily watch catches new dead links the morning they appear.

Related: redirect chains · orphan pages · broken outbound links

Questions

Keep reading