Every website collects broken links. It does not matter how carefully you publish, how disciplined your editorial team is, or how small the site happens to be. Pages get deleted, URLs get restructured, partners shut down their domains, and product lines disappear. The result is a slow, invisible decay that visitors notice before you do. A broken links checker exists to reverse that decay by finding every dead link before it costs you a reader, a customer or a ranking.
This guide explains what a broken links checker actually does, why link rot is far more widespread than most site owners assume, how to run a scan that catches everything, and how to turn a list of errors into a repeatable maintenance routine. Whether you manage a five-page portfolio or a 100,000-URL ecommerce catalogue, the principles are the same.
What a broken links checker actually does
At its core, a broken links checker is a crawler. It starts at a URL you give it, requests that page, extracts every link it finds, and then follows those links to see whether they resolve. For each link it records the HTTP status code returned by the server. A 200 means the page loaded. A 404 means the page was not found. A 410 means it was deliberately removed. A 5xx code means the server itself failed. A 301 or 302 means the request was redirected somewhere else.
The good checkers do more than read status codes. They read the content of the page as well, because a growing number of sites return a 200 for pages that are really errors. These are called soft 404s, and they are one of the sneakiest problems in technical SEO. The page tells the browser everything is fine while showing the visitor a “not found” message. Search engines treat these pages as low quality, and a checker that only looks at status codes will never flag them.
A modern checker also looks beyond plain hyperlinks. Images, scripts, stylesheets and documents such as PDFs are all links in the technical sense. If a hero image on your homepage returns a 404, the page still loads, but the visitor sees a broken icon and a gap in the layout. If a stylesheet fails to load, the page may render as unstyled text. Checking these assets is just as important as checking anchor tags.
How common is link rot, really?
Most site owners drastically underestimate how fast links die. Pew Research Center studied a decade of web pages and found that a quarter of pages that existed between 2013 and 2023 were no longer accessible by October 2023. Even more striking, roughly 8 percent of pages that were only a year old had already vanished. Ahrefs ran a separate study on backlinks and found that at least two-thirds of links pointing to websites over a nine-year period were dead.
Those numbers describe the entire web, but they apply to your site too. If you have been publishing for five years and have never audited your outbound links, a realistic expectation is that somewhere between 10 and 30 percent of them no longer work. Internal links fare better because you control both ends, but every redesign, migration or content pruning session leaves a trail of dead URLs behind.
Why broken links hurt more than you think
The obvious cost of a broken link is the visitor who clicks it and hits an error page. Some of those visitors will use the back button and try another link. Many will leave entirely. On a commercial site, a dead link on a pricing page or checkout flow is lost revenue, plain and simple.
The less obvious costs are the ones search engines impose. Three mechanisms matter here.
Crawl budget. Search engines allocate a finite number of requests to your site during each crawl. Every request spent on a URL that returns 404 is a request that was not spent discovering or refreshing a real page. On large sites this directly slows down how quickly new content gets indexed.
Link equity. Internal links pass ranking signals between pages. When a link points to a dead URL, the equity it would have passed simply evaporates. If you have a strong page linking to a deleted resource, that value is wasted instead of flowing to a live page that could rank.
Quality signals. A site full of dead ends looks unmaintained. Search engines pick up on user behaviour such as quick bounces from error pages, and a pattern of those bounces is not something you want attached to your domain.
Preparing for a full-site scan
Before you run a broken links checker, a little preparation makes the results far more useful.
First, decide on scope. Do you want to check only internal links, or external links as well? Internal-only scans are faster and produce a list you can fix entirely on your own. Including external links gives a more complete picture but adds noise from third-party sites that are temporarily down.
Second, make sure your XML sitemap is current. A checker that reads your sitemap first can discover pages that nothing links to, which are precisely the pages most likely to be forgotten and full of stale links. If your sitemap has not been regenerated since a migration, fix that first.
Third, think about exclusions. Tag archives, pagination, tracking parameters and search result pages generate thousands of near-duplicate URLs that waste crawl time. Most checkers let you exclude URL patterns such as /tag/* or anything containing ?utm_. Set those up before you start so the crawl spends its time on pages that matter.
Running the scan step by step
The workflow with a web-based checker is straightforward. Enter your domain, choose whether you want the entire site or a single page, tick the options you need, and start the crawl. A well-built tool streams results in real time, so you can start reviewing errors while the crawler is still working rather than waiting for a final report.
As results arrive, group them. A sensible first pass sorts by status code:
- 404 and 410 errors on internal pages are your top priority. These are pages on your own domain that you either deleted or renamed. Each one needs a redirect or a link update.
- Soft 404s are the second priority. These often indicate a template problem, such as a product page that shows “out of stock” with a 200 status, or a search page that returns an empty result with no error code.
- 5xx errors need investigation. Sometimes they are transient. Sometimes they reveal a script or database issue on a specific template that has been broken for months.
- Redirect chains are not errors, but a link that hops through three 301s before landing is slow and leaks equity at every hop. Update the link to point directly at the final destination.
- External 404s are last. Decide whether to update the link to a replacement resource, link to an archived copy, or remove the link entirely.
For every broken link, a good report tells you three things: the broken URL, the source page where the link lives, and the anchor text the link hides behind. That combination is what makes fixing fast. You open the source page, search for the anchor text, and change the link. Without the source page you would be hunting through your entire site for a needle.
Fixing what you find
There are only four ways to fix a broken link, and choosing the right one matters.
Update the URL. If the page moved and you know the new address, change the link. This is the cleanest fix because it avoids a redirect entirely.
Add a redirect. If many pages link to a URL that changed, or if external sites link to it, set up a 301 redirect from the old address to the new one. This preserves link equity from every source, including ones you cannot edit.
Replace the resource. If an external page is gone for good, look for an equivalent resource on the same site or a comparable one. Archived copies from services like the Wayback Machine are a reasonable fallback for reference links.
Remove the link. If none of the above applies, unlink the text. A plain phrase is better than a dead link.
For WordPress sites, this process can happen without leaving the dashboard. The Broken Link Checker plugin lists the exact post and anchor for every dead link and lets you edit or unlink from the report itself. That removes the tedious post-by-post hunting that used to make link audits an all-day job.
Turning a one-off scan into a routine
The most important thing to understand about broken links is that fixing them once does not solve the problem. New links rot every month. The sites you link to will change their URL structures. Your own team will delete pages. A single audit is a snapshot, not a cure.
The fix is scheduling. Scan monthly for a typical business site. Scan weekly if you publish frequently or link out heavily, because both behaviours increase the rate at which dead links accumulate. If your checker supports scheduled re-scans with email alerts, turn them on so new errors reach your inbox within a day of appearing rather than months later when a customer complains.
It also helps to build link checking into your publishing workflow. Before a new article goes live, scan the single page. It takes seconds and catches typos in URLs, links to draft pages that were never published, and pasted links that broke during editing.
What separates a good checker from a mediocre one
Not all checkers are equal, and the differences show up on real sites. The features that matter most in practice are these.
Sitemap-first crawling finds orphan pages that link-following alone would miss.
Soft-404 detection catches the errors that status codes hide.
SSL checks flag expired or misconfigured certificates that cause browsers to block pages entirely.
Asset checking extends the scan to images, scripts, stylesheets and documents.
Live streaming results let you start work immediately instead of waiting for the full crawl.
Shareable reports mean you can hand a client or developer a single link rather than exporting spreadsheets.
Exclusion patterns keep the crawl focused and the report free of noise.
Platform integrations such as a WordPress plugin or Chrome extension bring the results into the place where you actually fix things.
Broken Link Checker offers all of the above, with a free tier that scans 250 pages per crawl, two scans a day without an account, and 50 scans a month with a free account. For most small and medium sites that is enough to run a full monthly audit at no cost. Paid plans add scheduled monitoring, email alerts, AI-assisted link fixes and a professional WordPress plugin, and the same account covers a wider suite of SEO tools including rank tracking, an internal link optimizer, an SEO and AEO audit and an uptime monitor.
A quick checklist
If you take nothing else from this guide, take this five-step routine:
- Regenerate your XML sitemap so the crawler can find every page.
- Run a full-site scan with external links and assets included.
- Fix internal 404s and soft 404s first, then redirect chains, then external links.
- Scan every new page before it goes live.
- Schedule a monthly re-scan and act on alerts within a week.
Broken links are one of the few SEO problems that are entirely within your control, cheap to find, and fast to fix. The sites that stay healthy are not the ones that never break a link. They are the ones that check regularly and repair what they find.
