Link Audit Glossary and Frequently Asked Questions
Plain definitions for the terms you will meet in any crawl report, plus straight answers to the questions we hear most about broken links and redirects.
- 301 redirect
- A permanent redirect. It tells browsers and crawlers that the requested URL has moved for good and that the new URL should be treated as the canonical location, including inheriting the old URL's ranking signals.
- 302 redirect
- A temporary redirect. It tells clients to fetch a different URL for now while keeping the original URL as the one to remember. Search engines have said it passes ranking signals much as a 301 does, so the cost of using it for a permanent move is not lost equity but a canonical URL that fails to move, which delays consolidation onto the new address after a migration.
- 307 and 308 redirects
- Stricter versions of 302 and 301 respectively. They require the client to repeat the request with the same method, which matters for form submissions and API calls but rarely changes how a link audit treats them.
- 404 Not Found
- The status code a server returns when it has no resource at the requested URL. A 404 is the correct response for a URL that should not exist; it is a problem only when something still links to it.
- 410 Gone
- A status code stating that the resource existed once and has been intentionally removed with no replacement. Search engines typically drop a 410 URL from their index faster than a 404.
- 5xx server error
- Any status code in the 500 range, indicating that the server failed to handle the request. During a crawl these are often transient, so a good audit retries them before reporting them as broken.
- Anchor text
- The visible, clickable text of a link. Audit exports include it because it tells a developer which link on a page is broken when the page contains many links to similar destinations.
- Broken link
- A link whose destination does not return usable content, typically because the target responds with a 4xx or 5xx code, times out, or resolves through a chain that ends in an error.
- Canonical URL
- The URL a page declares, usually through a link tag in its head, as the preferred version of itself when the same content is reachable at several addresses. A canonical that points to a redirecting or missing URL is a frequent audit finding.
- Crawl budget
- The number of requests a search engine is willing to make to a site over a given period, shaped by server speed and the site's perceived value. Broken links and redirect hops consume budget that could have gone to useful pages.
- Crawl depth
- The number of link hops from a seed URL, usually the homepage, to a given page. Pages at a great depth are discovered and refreshed less often, which makes depth a useful column in an audit.
- Crawler
- Software that fetches a page, extracts the links it contains, and follows them to fetch further pages. Search engines run crawlers to discover content; link audit tools run them to test it.
- External link
- A link from a page on your site to a page on a different domain. Audits test external destinations for their status but do not follow links found on those destinations.
- Hop
- A single redirect step between a requested URL and the next URL the server sends the client to. A redirect chain is measured in hops; the ideal number is zero, and the acceptable maximum is one.
- Hreflang
- An attribute that tells search engines which language or regional version of a page to show to which audience. Hreflang references that point at redirecting or missing URLs break the relationship and show up in a thorough audit.
- Internal link
- A link from one page on your site to another page on the same site. Internal links are how crawlers discover most of a site's pages and how importance flows between them.
- Link equity
- The informal name for the ranking value that a link passes from one page to another. A link to a broken URL passes none of that value. Whether a redirect chain sheds some of it at each hop is a common SEO belief rather than a documented mechanism.
- Link rot
- The gradual decay of links over time as the pages they point to are moved, renamed, or deleted. Outbound links to other sites rot faster than internal ones because you do not control the destination.
- Log file analysis
- Reading a web server's access logs to see which URLs crawlers and visitors actually requested and what responses they received. It reveals 404s and chains that a crawl from the homepage may never reach.
- Nofollow
- A link attribute that asks search engines not to pass ranking signals through the link. Nofollow links can still be broken, and an audit should test them like any other.
- Orphan page
- A page that exists on the site but has no internal links pointing to it. Orphans cannot be found by following links, so they are only detected by comparing the crawl against sitemaps, logs, or analytics.
- Redirect chain
- A sequence of two or more redirects between a requested URL and the page that finally loads. Chains add latency, waste a crawl request at each hop, and can exceed the number of hops a crawler is willing to follow. The idea that they also lose a little equity at each hop is a common SEO belief rather than a documented mechanism.
- Redirect loop
- A chain in which some URL redirects, directly or through other hops, back to itself. The page never loads, browsers show an error, and crawlers abandon the URL.
- Rendering
- Executing a page's JavaScript during a crawl so that links added by scripts become visible. Rendering finds links that plain HTML parsing misses, at a significant cost in crawl time.
- Robots.txt
- A file at the root of a site that tells crawlers which paths they may or may not fetch. An audit crawler should respect it by default, and a path blocked here can explain why a page appears missing from a crawl.
- Seed URL
- A starting point given to a crawler, from which it begins extracting and following links. The homepage and sitemap URLs are the usual seeds; adding more seeds improves coverage.
- Soft 404
- A page that returns a 200 success code but contains no real content, such as an empty listing or a generic error message in the site template. Search engines treat it as missing, but only after spending a request to find out.
- Status code
- The three-digit number a server sends with every response to say what happened to the request. Link audits are largely organized around these codes: 2xx success, 3xx redirect, 4xx client error, 5xx server error.
- URL normalization
- The process of reducing URL variants such as trailing slashes, letter case, tracking parameters, and protocol differences to a single canonical form so a crawler does not count the same page several times.
- XML sitemap
- A machine-readable file listing the URLs a site wants crawled. Comparing the sitemap against the crawl results is one of the fastest ways to find orphan pages and URLs that redirect or error.
Questions people ask
What exactly does a link audit check?
It requests every URL the site exposes and records the response: the status code, any redirect hops, the final destination, and the page that contained the link. From that data it reports broken links, redirect chains and loops, soft 404s, server errors, and pages that exist but are not linked from anywhere.
Do broken links really affect search rankings?
Not as a direct penalty, but through their effects. Broken internal links cut off pages from discovery and stop ranking signals from flowing. Broken outbound links make a page look unmaintained to visitors. Redirect chains waste crawl requests and add latency for every visitor. Individually small, these effects add up on a site that never audits.
How many redirect hops are acceptable?
One. A single redirect from an old URL to its replacement is normal and harmless. Two or more hops is a chain worth collapsing so that the first redirect points straight at the final destination. The best outcome is to update the source link so no redirect is needed at all.
Should I redirect every 404 to the homepage?
No. Blanket redirects to the homepage create soft 404s, because the destination has nothing to do with what the visitor asked for. Redirect a removed page to a genuinely equivalent page if one exists. If none does, let it return 404, or 410 if you want to signal that it is gone for good.
What is the difference between a 404 and a soft 404?
A real 404 is an honest response: the server says the page does not exist. A soft 404 is a page that says or shows nothing useful while returning a 200 success code. Search engines eventually treat both as missing, but the soft version wastes crawl effort and hides the problem from simple status checks.
Why did my crawl find fewer pages than my sitemap lists?
Usually because some pages are orphans with no internal links, are blocked by robots.txt, sit behind a login, or are only linked through JavaScript the crawler did not render. Compare the two lists; the pages in the sitemap but absent from the crawl are the ones to investigate first.
Can a link audit slow down or break my live site?
A crawl that fires requests too quickly can strain a small server and produce 5xx errors for real visitors. Reputable audit tools throttle their crawl rate and identify themselves with a user agent. If you are cautious, crawl during a low-traffic window or have the crawler's address allowlisted with a rate limit.
How long does a full link audit take?
It depends on page count, server response time, and crawl rate. A few thousand pages typically finish within an hour at a polite rate. Very large sites with hundreds of thousands of URLs can take many hours or need to be crawled in segments, especially if JavaScript rendering is enabled.
Should external links be included in the audit?
Yes. Outbound links rot faster than internal ones because other sites move and delete content without telling you. The audit should test each external destination's status but stop there, without crawling the external site further.
What should I fix first after an audit?
Group the findings by cause, then rank by impact. Template-level issues that affect every page, broken links in main navigation, chains on high-traffic landing pages, and soft 404 patterns across whole sections come first. A single 404 on an old, unlinked post can wait.
Do I need to audit links before a site launch?
It is the single highest-value moment to do so. Crawl the staging site fully, test every old URL against the new redirect rules, and verify navigation, sitemaps, canonicals, and hreflang before go-live. Then crawl production within the first days to confirm the live rules behave as tested.
How do I hand audit results to a developer?
Give them one row per distinct problem rather than one row per broken instance, with the source template or page, the anchor text, the target URL, the observed status and final destination, the recommended fix, and the number of instances affected. Attach the raw export separately for reference.
Is it ever correct to leave a 404 in place?
Yes. A URL that never should have existed, a retired page with no equivalent, or a URL that nothing links to can legitimately return 404 or 410. The problem is not the 404 itself but any link, sitemap entry, or canonical tag that still points at it.