An internal linking audit checks how pages on your own site connect to one another. It can reveal broken destinations, pages a crawler cannot reach through links, weak navigation paths, and anchors that give readers little context. The aim is not to maximize link counts. It is to make important pages reachable through useful, understandable paths.
What is an internal linking audit?
An internal linking audit is a review of links between pages on one website, including their source, destination, anchor text, status, and place in the site structure. It combines crawl data with business and user context because a technical link count alone cannot show whether a page matters or whether the path is useful.
Google Search Central guidance says Google uses links to find pages and understand relevance, and recommends crawlable links with descriptive anchor text. This supports checking link implementation and context. It does not mean a specific number of internal links guarantees rankings.
How do you check internal links?
Start with a crawl of the canonical site version, then compare its URL set with other records such as XML sitemaps, analytics landing pages, and a known CMS export. A crawler only sees the site paths available to its configuration; a single crawl may miss pages behind authentication, filters, JavaScript behavior, or an incomplete starting set.
1. Define scope and collect URL evidence
Record the domain, subdomains, protocol, locale paths, and page types in scope. Decide how redirects, canonical tags, parameters, and staging environments should be handled before interpreting counts.
Crawl from representative entry pages and include the XML sitemap as a discovery source if the tool supports it. Then compare crawl URLs with sitemaps and analytics. A URL found in analytics but absent from the crawl is an orphan candidate, not proof of an orphan: it may be linked only through a path your crawl did not include.
2. Find broken, redirected, or blocked destinations
Export internal links and their response status. Separate links returning errors from links that resolve through redirects. Confirm whether a destination is intentionally retired, temporarily unavailable, or incorrectly linked before editing.
Prioritize broken links on high-use pages, navigation, and conversion paths. Do not automatically replace every redirect: first check that the final destination is relevant and intended. A technically successful response can still lead to the wrong page.
3. Identify orphan candidates and pages with weak paths
An orphan page is a page with no internal links pointing to it from within the defined site set. The exact result depends on crawl scope and access. Compare crawl data with the sitemap, CMS inventory, and analytics landing pages before declaring a URL orphaned. Google describes sitemaps as a way to give Search information about site pages and their relationships, but they should be treated as a discovery source rather than proof of an internal link path.
For reachable pages, examine click depth and incoming internal-link counts as clues. Neither metric is a standalone quality score. A deep support article may be appropriate; a critical service page buried many clicks away may warrant a clearer route.
4. Review anchor context and link placement
Anchor text is the visible link text. Google recommends descriptive anchor text that helps people understand the destination. Review links such as “click here,” generic repeated anchors, and text that does not match the destination. Also inspect the surrounding sentence: a clear anchor in an irrelevant paragraph still creates a poor reader path.
Screaming Frog's tutorial describes using crawl reports to inspect link counts, discover opportunities, and locate non-descriptive anchor text. Treat its tool-specific workflow as one implementation option, not the only way to audit a site.
5. Build a prioritized fix queue
Do not sort only by total inlinks. Give each candidate a short assessment of:
Criterion | Question |
|---|---|
Page importance | Does the destination support a key buyer task or site function? |
User path | Does the source page help the reader reach the destination at this point? |
Severity | Is the destination broken, inaccessible, misleading, or merely weakly linked? |
Evidence | Do crawl, analytics, sitemap, and CMS records agree? |
Effort | Can the fix be made safely without changing unrelated navigation? |
A simple priority queue can use “fix now,” “investigate,” and “monitor.” For example, a broken link from a high-traffic product page to a current pricing resource may merit immediate correction. A page with few incoming links but no clear audience role may need a content or retirement decision before adding more links.
For a reproducible queue, keep the observed issue separate from the proposed remedy. “Four internal links return 404” is a crawl observation; “replace them with the new overview page” is a recommendation that still needs a relevance check. Record the evidence source, affected URLs, expected user benefit, risk, owner, and verification step. This makes review easier when a fix touches navigation or a large set of templates.
A hypothetical example: an audit finds a broken link on an important pricing explainer and also a deep editorial page with few incoming links. The broken commercial path may rank higher because it blocks a useful next step; the deep article should first be checked for audience value and intended discoverability. Neither decision follows from the count alone.
6. Apply and verify fixes
For each fix, preserve the source page, destination, proposed anchor or placement, reason, owner, and status. Make one small change at a time where possible. Then open the page, confirm that the link is visible and points to the intended URL, and rerun a crawl or targeted check.
If the destination is unpublished or its canonical URL is unknown, do not add a placeholder link. Keep the proposal private until the destination is confirmed. Likewise, avoid adding links solely to raise an inlink count; each link should help a reader move to a relevant next step.
How often should you audit internal links?
The right cadence depends on how often the site changes and how critical its navigation is. Recheck after migrations, major information-architecture changes, bulk content updates, or recurring crawl errors. For a stable site, periodic sampling plus monitoring of broken links may be more useful than repeating a full audit on an arbitrary schedule.
Keep a baseline of scope, crawl settings, exclusions, and date. Without those details, changes in counts may reflect a different crawl rather than an actual site change.
Frequently Asked Questions
How do you check internal linking?
Crawl the site, export source and destination URLs with status and anchor text, then reconcile the crawl with sitemaps, CMS records, and analytics. Inspect suspicious cases manually because crawler coverage and configuration can affect the results.
What should an internal linking audit checklist include?
Check crawl scope, broken and redirected destinations, orphan candidates, click depth, incoming links, anchor context, and high-value user paths. Record the evidence and verify each repair after implementation rather than treating a count threshold as a universal standard.
How Nortav Can Help
When an audit report flags dozens of weak links but gives no clear order, I can help distinguish critical reader paths from low-impact counts and set a defensible repair queue. Our approach starts with understanding current search visibility, identifying gaps against competitors, and developing a strategy focused on qualified traffic, leads, and measurable business value. Book a Discovery Call →
Sources
Google Search Central, “Make your links crawlable.” https://developers.google.com/search/docs/crawling-indexing/links-crawlable
Screaming Frog, “Internal Linking Audit With the SEO Spider.” https://www.screamingfrog.co.uk/seo-spider/tutorials/internal-linking-audit-with-the-seo-spider/
Google Search Central, “Learn about sitemaps.” https://developers.google.com/search/docs/crawling-indexing/sitemaps/overview
Cached Serper US/en SERP for “internal linking audit,” collected 2026-10-18, used for PAA and related-search provenance only.
