Home / Glossary
Glossary and FAQ

Broken Link Monitoring Glossary and FAQ Hub

Reviewed for 2026, updated September 8, 2026.

Plain-language definitions of the terms you will meet when finding, fixing, and monitoring broken links, followed by answers to the questions site owners ask most.

345ABCDEHILMNORSTUX
301 redirect
An HTTP response that tells browsers and search engines a URL has moved permanently to a new address. It is the standard way to preserve traffic and ranking signals when a page changes location, and the new URL is treated as the canonical one over time.
302 redirect
An HTTP response indicating a temporary move. Search engines treat it as a weak canonicalization signal, so the original URL tends to stay indexed instead of being replaced by the destination. It is often used by mistake when a permanent 301 was intended.
404 Not Found
The HTTP status code a server returns when the requested URL does not exist. It is the most common form of broken link. A well-designed 404 page keeps the visitor on the site by offering navigation, search, and likely alternatives while still returning the correct status code.
410 Gone
A status code that says the resource existed but has been intentionally and permanently removed. Compared to a 404, it gives search engines a clearer signal to drop the URL from their index.
5xx server error
Any status code in the 500 range, such as 500, 502, or 503, indicating the server failed to fulfill a valid request. A link that returns a 5xx is broken from the visitor's point of view even though the page may exist, and monitoring tools usually retry before reporting it because these errors are often transient.
Anchor text
The visible, clickable text of a link. It tells visitors and search engines what to expect at the destination, which is why a broken link with descriptive anchor text is doubly frustrating: it made a specific promise and then failed to keep it.
A link from another website to a page on yours. Backlinks are a major ranking factor, and when the page they point to is removed or moved without a redirect, the value they carry is lost. Preserving backlink destinations is a central concern during migrations.
Canonical URL
The preferred address for a page when several URLs serve the same or very similar content, declared with a canonical tag. Crawlers use it to avoid treating duplicates as separate pages, and a canonical tag pointing to a broken or redirected URL is a common hidden problem.
Content drift
The gradual change of a page's content after it has been linked, until the link no longer points to what the linking page described. The link still resolves, so status-code checks do not catch it. It is the quieter cousin of link rot and mostly affects citations and references.
Crawl budget
The number of URLs a search engine crawler is willing to request from a site over a given period. Requests spent on broken URLs, redirect chains, and duplicate pages reduce what remains for content that should be discovered and refreshed.
Crawler
Software that visits a starting URL, extracts the links on the page, and follows them to discover and check more pages. Search engines run crawlers to build their index; broken link tools run crawlers to record the status of every link they find. Also called a spider or bot.
DNS failure
An error that occurs when the domain name in a URL cannot be resolved to a server address, typically because the domain has expired, been misconfigured, or never existed. A link with a DNS failure is broken at a deeper level than a 404, since no server can even be reached.
A link from your site to a page on a different domain, also called an outbound link. You do not control the destination, so external links break over time regardless of how well your own site is maintained, and they are the main source of link rot in older content.
HTTP status code
A three-digit number returned with every web response that summarizes the result. Codes in the 200 range mean success, 300 range means redirection, 400 range means a client-side problem such as a missing page, and 500 range means a server-side failure. Link checkers rely on these codes as the first signal of a broken link.
A link between two pages on the same site. Internal links carry visitors and ranking signals through the site, and because you control both ends, an internal broken link is entirely your responsibility and can always be fixed completely.
The informal name for the ranking value that passes from one page to another through a link. A broken link stops that flow at a dead end, and a redirect chain puts it at risk because crawlers follow only a limited number of hops before giving up. Repairing broken internal links restores equity to the pages that were losing it.
The tendency of links to stop working over time as the pages they point to are moved, removed, or replaced, and as domains expire. It affects every long-lived site and is the reason a page that was perfect when published needs periodic checking.
Meta refresh
An HTML instruction that sends the browser to another URL after a delay, sometimes used as a makeshift redirect. It is slower than a server-side redirect, is treated inconsistently by search engines, and can hide a redirect chain from simple status-code checks.
Nofollow
A link attribute that asks search engines not to pass ranking signals through the link. A nofollow link that is broken is still a broken link for the visitor; the attribute changes SEO treatment, not the user experience.
Orphan page
A page that exists on the site but has no internal links pointing to it. Crawlers that start from the home page cannot find orphan pages, so they are discovered by comparing a crawl against the sitemap, server logs, or the content management system's own list of pages.
Rate limiting
A server's practice of restricting how many requests a single client can make in a period of time. Crawlers that check external links too aggressively get rate-limited or blocked, which produces false reports of broken links. Responsible checkers space out requests and retry before concluding a page is gone.
Redirect chain
A sequence in which a URL redirects to a second URL, which redirects to a third, and so on, before reaching the final page. Chains add latency for visitors, waste crawl budget, and can run past the number of hops a crawler is willing to follow, leaving the final page unreached. The fix is to point every old URL directly at the final destination.
Redirect loop
A redirect chain that circles back to a URL already visited, so the browser never reaches a page and eventually gives up with an error. Loops are usually caused by conflicting redirect rules and are a hard failure for both visitors and crawlers.
Redirect map
A document, usually a spreadsheet or configuration file, that pairs each old URL with its new destination during a migration or restructure. A complete, one-to-one redirect map is the single most effective protection against a wave of broken links after a site change.
Robots.txt
A text file at the root of a domain that tells crawlers which paths they may or may not request. Link checkers should respect it when crawling other people's sites, and a misconfigured robots.txt on your own site can hide whole sections from both search engines and your monitoring.
Soft 404
A page that displays a not-found or empty message to the visitor but returns a 200 success status code. Because the code says everything is fine, simple link checkers miss soft 404s, and search engines may keep the useless page indexed. Detecting them requires looking at the page content, not just the status.
Timeout
A failure that occurs when a server does not respond within the time a client is willing to wait. Timeouts can indicate a dead host or merely a slow one, so monitoring tools typically retry a timed-out link before reporting it as broken.
User agent
The identifier a browser or crawler sends with each request to say what it is. Some servers respond differently to crawlers than to browsers, blocking or redirecting them, which is why a link can appear broken in a checker but work fine when clicked by a person.
XML sitemap
A machine-readable file listing the URLs a site wants search engines to know about. It is a useful reference for link monitoring because comparing it to a crawl reveals orphan pages, and any URL in the sitemap that returns an error is a high-priority fix.

Questions people ask

What is broken link monitoring?

Broken link monitoring is the practice of regularly checking every link on a website, both internal and outbound, to confirm that each one still reaches a working page. Instead of a one-time audit, monitoring runs on a schedule, keeps a history, and reports new failures as they appear so they can be fixed while they are still cheap to fix.

Why do links break in the first place?

Pages get moved, renamed, or deleted without a redirect. Domains expire or change hands. Content management systems change their URL structure after upgrades or migrations. Editors mistype an address. And on the outbound side, other people's sites change for all the same reasons, which is entirely outside your control. Given enough time, every long-lived site accumulates dead links.

How many broken links is too many?

There is no magic threshold, and a small number of 404s on old content is normal for any site. The concern is where they are and what share of the site they affect. A single broken link in the main navigation or a checkout flow is worse than dozens in archived posts. As a rule of thumb, aim for zero broken internal links on important pages and a steady process for the rest.

Should I redirect every broken URL?

No. Redirect a URL when there is a genuinely equivalent destination for the visitor and for search engines. If the page was removed on purpose and nothing replaces it, a clear 404 or 410 with a helpful error page is the honest answer. Redirecting everything to the home page confuses visitors and is treated by search engines much like a soft 404.

What should I do about broken external links?

First decide whether the reference still matters. If it does, look for the same content at a new address on the same site, or link to an archived copy so readers can still see what was cited. If the reference no longer matters, remove the link or rewrite the sentence so it stands on its own. Do not leave a dead outbound link in place on the assumption that nobody clicks it.

Do broken links affect page speed or Core Web Vitals?

A broken hyperlink in body text does not slow the page that contains it. Broken references to resources the page loads, such as scripts, stylesheets, fonts, and images, are a different story: they can delay rendering, cause layout shifts when an image fails, and add failed requests to the page load. Monitoring should cover those resource links as well as navigational ones.

Can broken links be found without crawling the whole site?

Partially. Server logs and analytics show which URLs visitors and search crawlers are actually requesting and which return errors, and search engine webmaster reports list the failures their crawlers encountered. Those sources are valuable, but they only reveal problems that someone has already hit. A crawl is the only way to find broken links proactively, before a visitor does.

Why does a link checker report a link as broken when it works in my browser?

The destination server may treat automated clients differently from browsers, returning an error, a block page, or a redirect based on the user agent or request rate. Some sites also require cookies, JavaScript, or a login. Good checkers retry, identify themselves honestly, and flag these cases separately so you can verify them by hand rather than treating every report as a confirmed failure.

How long should redirects stay in place after a migration?

As long as anything still points at the old URLs, which in practice means years. Search engines keep requesting old addresses long after a move, and backlinks from other sites are never updated. Removing redirects after a few months quietly reintroduces the very broken links the migration was supposed to avoid. Review them periodically for chains, but keep them.

What is the difference between a broken link audit and monitoring?

An audit is a one-time snapshot: crawl the site, list the failures, fix them. Monitoring repeats that process on a schedule, remembers the previous results, and reports only what has changed. An audit tells you where the site stands today; monitoring keeps it that way without depending on someone remembering to check again.

Who on the team should own broken links?

It usually takes more than one person, because fixes fall into different areas. Developers own redirect rules, templates, and shared components. Content editors own links inside articles and pages. Marketing owns campaign landing pages and links in emails. The most reliable setup names one person to receive alerts and triage them, and routes each fix to whoever owns that part of the site.

Does a custom 404 page fix the SEO problem of a broken link?

No. A good 404 page reduces the damage to the visitor by offering a way forward, but the link is still broken, the crawler still hits a dead end, and any ranking signals still stop there. Treat the error page as a safety net and fix the links that send people to it.