How do broken links affect your website and your SEO?
Where broken links come from, what they do to the people who click them and to the search engine crawling your site, how internal and external ones differ, and where to start fixing them.
By Javier Tenjo, Full stack developer
On this page
A broken link is a link that leads to a page that no longer exists. Whoever clicks it expects to find something and gets an error instead, nearly always the familiar 404. It is one of the most common problems on any site that has been around for a few years, and one of the least visible to its owner: you do not click your own links every day, but your visitors do.
You do not need to neglect anything to end up with them. Time passing is enough.
Where broken links come from
They are rarely a typo. Usually the link was fine when it was published and what surrounded it changed:
- A page was deleted or unpublished. A discontinued product, a service you no longer offer, an offer that expired. The links pointing to it from other pages are still there.
- A URL changed without redirecting the old one. Fixing an article’s name or reorganising the menu changes the address, and everything that linked to the previous one is left hanging.
- A migration or redesign. Changing platforms usually changes the whole URL structure. Without a redirect plan, the links in older content point to addresses that no longer respond.
- The other site changed. Links to other people’s sites break without you doing anything: the article you cited moved, the company closed, the domain expired. This is what is known as link rot, and over the years it affects a noticeable share of any site’s outbound links.
- Typos, this time for real. An extra slash, a swapped letter, a link copied from the staging environment with its temporary domain.
What a broken link does to your visitor
This is the most direct effect and the easiest to underestimate. Someone clicking a link is in the middle of a decision: they want to see the price, read about a service in detail or download a document. An error page interrupts that path just as it was moving forward.
Some people use the site search or go back. Many simply leave. And a broken link in the menu or on the main button of a service page lets go of visits that had already made it to you.
It also signals neglect. A site with links that lead nowhere gives the impression nobody looks after it, and that weighs on whether someone trusts you as a supplier.
What it does to the search engine
Precision helps here, because exaggerated ideas are going around. Google has explained that 404 errors are a normal part of the web and that they do not penalise the whole site for having them. A page that no longer exists should return 404 or 410: that is the correct answer.
The problem is not the 404 itself but what is lost with the broken link:
- Wasted crawling. Every broken link is a request the search engine makes for nothing. On a small site it hardly matters; on one with thousands of pages, it is crawl time not spent on the pages that do count.
- Signal that goes nowhere. Internal links are how you tell Google which pages matter and what they are about. A link pointing to a 404 passes that signal to nobody.
- Lost external authority. If other sites linked to a page of yours that no longer exists, that endorsement is lost. It is the costliest case, because those links are hard to earn. This is covered in backlinks, authority and internal linking.
Internal versus external links
They are two different problems and they are fixed in different ways.
| Internal links | External links | |
|---|---|---|
| Point to | Your own site | Third-party sites |
| Why they break | Your changes: deletions, new URLs, migrations | Other people’s changes you do not control |
| How to fix | Correct the link or redirect the old URL | Update the link, find another source or remove it |
| Impact | Crawling, internal signal and experience | Mainly experience and credibility |
There is a third category that is not yours but does affect you: links from other sites to pages of yours that no longer exist. You cannot edit those links, but you can redirect the old URL to the equivalent current page, and that recovers both the visit and the signal.
Where to start fixing
If a review returns dozens of broken links, there is no need to work through them alphabetically. This order usually gives the best return per hour:
- Links in the menu, header and footer. They repeat on every page, so one broken link there is hundreds of broken links for the search engine and a stumble for every visitor.
- Conversion pages. Services, pricing, contact, product pages. A broken link here costs customers.
- Old pages with inbound links or traffic. If a URL that disappeared is still getting visits or links from other sites, redirect it with a 301 to its closest equivalent.
- Internal links in body content. Articles and pages linking to others of yours that changed address.
- External links. The least urgent for rankings, but worth reviewing from time to time, especially in content that still gets visits.
One detail: when an internal link points to a URL that redirects, correct the link so it points to the final destination. The redirect is there for people arriving from outside; your own links should not depend on it. Why chains of hops are a problem is explained in 301 vs 302 redirects.
False positives
They exist, and it is worth knowing before deleting anything. Some servers block automated requests and answer 403 or 429 to a tool even though the page works perfectly well in a browser. Others are slow to respond because they are having a bad day, or return a temporary server error (the 5xx codes) that clears up on its own.
Before removing an external link flagged as broken, open it by hand. And for server errors, check again after a while.
How often to review
It depends on how much your site changes. A review after every migration, redesign or content reorganisation is essential. Beyond that, a periodic check of the main pages (every quarter, say) catches the external links that have broken on their own. The sitemap is a good starting point for knowing which pages to review.
Checking yours
The broken link checker reads a page, extracts all its links and checks one by one what each of them returns. It groups them into broken, server errors, no response, redirects, slow and servers that block robots, so you know which are genuinely a problem and which are worth checking by hand. It works page by page, so the practical approach is to start with the ones that matter most: the home page, the service pages and the contact page.
For links that redirect, the redirect checker shows you the full chain and the final destination you should be linking to.
Broken links cannot be avoided entirely, because the web changes on its own. What you can do is find them before your customers do, fix the costliest ones first and redirect every URL you retire. That way the problem stays routine maintenance.