Teardown
Broken internal links: usually one mistake, repeated on every page
Six in ten sites I audited linked to at least one of their own pages that no longer works.
On this page
26 of the 42 startup sites I audited linked to at least one of their own pages that no longer worked. Across all of them, 574 of the 9,705 internal link targets I checked were broken.
These were not flukes of the crawl. When I fetched a sample again on 20 September, 47 of 48 were still broken.
What I found
42 startup sites, 14,023 pages, crawled 15 to 18 September 2026
- Sites affected
- 26 of 42 (62%)
- Typical size
- 7 broken targets on the median affected site
- Across all sites
- 574 of 9,705 link targets checked were broken
- Re-checked live
- 47 of 48 sampled links still broken on 20 September
Why one dead link shows up everywhere
Most affected sites had only a few broken targets, and that small number is misleading. A link in the header, the footer or a sidebar sits on every page, so one dead link there is one mistake shown to every visitor.
The cause is nearly always a page that was removed or renamed while something still pointed at it: an old pricing page, a retired feature, a blog post merged into another, a documentation page that moved.
What Google does with a page that returns an error
Google's documentation on HTTP status codes is short on this:
In the case of Google Search, Google doesn't index URLs that return a 4xx status code, and URLs that are already indexed and return a 4xx status code are removed from the index.
Google Search Central, HTTP status codes
So a broken link points at a page that cannot appear in search. If that page mattered, it has gone from search as well as from the visitor's path. If it did not, the cost is mostly a person who clicked and hit a dead end.
How to fix it
Find where the link lives before fixing it. If it is in a shared part of the site, such as the header or footer, one edit fixes every page.
If it points at a page that moved, update the link to the new address, and add a redirect from the old one for anyone else still using it. If the page is gone for good, remove the link.
Check your own site, no code needed
- Start with the links that repeat: click every link in your header, footer and main navigation.
- Once Search Console is set up, open the Pages report and look under Not found (404).
- Paste one of those addresses into URL Inspection. The Referring page field shows a page Google may have found the link on. It is not always filled in, so an empty field does not mean nothing links there.
What to ask your developer
- Is this link in a shared template, like the header or footer?
- When a page is removed, does anything check for links still pointing at it?
What I could not tell from outside
Broken means the last response was an error when fetched from India with the crawler I used. A page behind a login, or blocked by region, can look broken from outside while working for its real users. The individual audits checked for this; this summary does not.
These are counts on 42 startup sites I audited from the outside, mostly developer tools and SaaS. They are not an industry average, and nothing here measured whether any of these sites performed worse in search because of what I found.
Want to know which of these your site has?
Send a URL. Three to five verified findings come back within two working days, free, each with a way to check it yourself.