You've published the pages, built a navigation menu and added a few related-article modules. Yet important service, product or guide pages still sit deep in the site, receive almost no internal links, or can only be found through an XML sitemap. That's the situation an internal linking audit is designed to fix.
A proper audit maps how pages connect, how far they sit from the homepage, where links are broken or misleading, and whether your most commercially important pages receive enough support. It also checks whether visitors can complete tasks without hitting dead ends. Internal linking is therefore both an SEO control and a usability control, not a cosmetic exercise.
Table of Contents
- What an Internal Linking Audit Actually Fixes
- Preparing Your Crawl and Link Map the Right Way
- How to Read Your Internal Link Data and Spot Problems
- Prioritising Fixes by Business Value Not Just Link Count
- Remedial Actions That Strengthen Your Site Architecture
- Measuring Impact and Reporting Your Internal Linking Audit
What an Internal Linking Audit Actually Fixes
An internal linking audit identifies weaknesses in the routes connecting pages on your website. It shows whether search engines can discover important URLs, whether internal authority is concentrated in the right places, and whether users can move logically from an informational page to a relevant product, service or enquiry page.
The output shouldn't be a spreadsheet containing thousands of URLs with no order. It should be a prioritised list of problems such as:
- Orphaned pages: URLs found in an XML sitemap, analytics or Search Console but not through normal internal links.
- Excessive crawl depth: Important pages buried more than three clicks from the homepage.
- Broken routes: Internal links leading to 4xx pages, redirects or redirect chains.
- Weak anchors: Links labelled “click here”, “read more” or other wording that doesn't explain the destination.
- Misdirected support: Blog content receiving substantial internal attention while money pages remain underlinked.
Google's internal linking documentation makes the discovery role clear, while the Screaming Frog internal linking audit workflow recommends exporting crawl depth, inlinks, outlinks, status codes and anchor text so these issues can be assessed systematically.
The SEO outcome is only half the job
Internal links help search engines understand relationships between pages and infer which URLs matter within the site. They also shape the user's next action. A guide about choosing a Shopify platform should lead naturally to an ecommerce service page or relevant implementation guide, not leave the visitor at the bottom of an article with nowhere useful to go.
Accessibility adds an important test. UK government content guidance says link text should describe the destination clearly, and the GOV.UK guidance on writing links treats link purpose as part of good content design. DWP accessibility testing guidance also treats broken internal links as a failure condition, which means a link can be poor even when it passes some narrow SEO checks.
Practical rule: Don't ask only whether a page has enough links. Ask whether the right people and crawlers can reach the right destination at the right point in the journey.
The scale of structural defects is why this belongs in routine governance. The Government Digital Service reported 29,787 accessibility issues across monitored public-sector websites and mobile apps in its 2022–2024 review (GOV.UK accessibility monitoring report). That figure isn't an internal-linking statistic, but it demonstrates how small structural weaknesses can multiply across large estates.
The same principle applies to a content-heavy commercial website. A small error rate across thousands of URLs can create hundreds or thousands of affected pages. An audit is worthwhile when navigation has changed, a site has been migrated, publishing has accelerated, important pages aren't being indexed, or organic performance doesn't reflect the value of the content being produced.
For teams considering machine-assisted recommendations, an internal linking automation guide can help explain where automation fits. I'd still treat automated suggestions as a review queue, not an instruction to link every matching phrase.
Preparing Your Crawl and Link Map the Right Way
A crawler is only as useful as its scope and starting conditions. Before analysing links, define what the audit must include and what should be excluded. A blog-only review needs a different crawl from an ecommerce site with product variants, filters, account areas and multiple subdomains.
Start by writing down the boundaries. Include the canonical site version, relevant subdomains and any high-value URL sets that aren't linked from the homepage. Exclude staging environments, internal search results, user account pages and irrelevant parameter combinations where they don't form part of the indexable site architecture.
Build a crawl that reflects the real website
Use the homepage as the primary start URL for a normal architecture audit. Add key landing pages or known sitemap URLs when the site contains sections that the homepage doesn't link to. This prevents the crawl from confusing “not linked from the chosen start point” with “orphaned across the whole website”.
In Screaming Frog, Sitebulb or a comparable crawler, export at least:
- Source and target URLs: Which page links to which destination.
- Crawl depth: The number of hops from the start URL.
- Status codes: Whether each destination returns a successful response, redirect or error.
- Inlinks and outlinks: The number and source of internal connections.
- Anchor text: The visible wording used for each link.
- Indexability signals: Canonical, noindex and robots directives where available.
The Sitebulb explanation of crawl depth and orphan-page auditing is useful here because it distinguishes between pages discovered through crawling and URLs found in supplementary sources. Sitebulb compares crawl data with Google Analytics, Google Search Console, XML sitemaps and optional URL lists. A URL appearing only in those sources isn't necessarily useless, but it is absent from the internal link map and deserves investigation.
Classify pages before judging them
Add a page-type column before you sort the export. Useful categories include service pages, product pages, category pages, blog posts, guides, case studies, location pages and documents.
Then add business context. Mark which pages generate enquiries, support sales, target strategic topics or act as important hubs. Without this classification, a crawler will treat a privacy policy and a high-value service page as equivalent rows.
You should also compare the crawl against the XML sitemap. The UK internal linking audit workflow from Link Building Journal recommends this comparison because sitemap URLs absent from the internal crawl may be orphaned. The comparison doesn't prove that every URL needs more links, but it gives you a reliable review set.
Crawl budget deserves proportional attention. The Sibley Digital guide to crawl budget is relevant when a website is large, frequently changing or producing substantial URL variation. Smaller sites still need clean internal links, but they generally don't need complex crawl-capacity engineering.
How to Read Your Internal Link Data and Spot Problems
The raw export becomes useful when you group rows into patterns. Start with destinations, not sources. Sort by unique inlinks, then compare each page's support with its business role and crawl depth.
An important service page with one route from an obscure article is more concerning than a low-value legal document with the same count. Likewise, a page with many inlinks can still be poorly supported if every link comes from a footer, a repetitive template or an irrelevant section.
Find orphans through comparison
A crawl only sees what it can reach from its starting URLs. That makes source comparison essential. Export URLs from Google Analytics, Google Search Console, your XML sitemap and any known CMS or product feed, then compare them with the crawler's discovered URL list.
Sitebulb describes an orphan page as a URL found through those other sources but not through internal links. The definition is operationally useful because it separates a genuine architecture problem from a crawler configuration mistake.
Investigate each candidate:
- Confirm the URL should exist: Remove retired, duplicate or deliberately excluded URLs from the action list.
- Check indexability: A noindex document or private resource may not need internal support.
- Assess value: A commercial page, key guide or category deserves more attention than a temporary document.
- Choose a relevant donor: Link from a page with genuine topical and user relevance, not merely the page with the highest link count.
Read depth as a structural signal
Crawl depth measures how many hops the crawler follows from the start URL. UK guidance from SEO Strategy on site architecture recommends keeping important content within three clicks of the homepage. The same principle appears in the Screaming Frog audit guidance.
Pages at four or more clicks aren't automatically worthless. A specialist resource may naturally sit deeper than a core service page. The problem appears when a page is commercially important, receives little internal support and sits deep because no one has designed a route to it.
Check for broken destinations and redirect chains at the same time. A link to a redirected URL may still work for a visitor, but it adds an unnecessary step for crawlers and creates maintenance debt. A link to a broken URL is more serious because it interrupts discovery and task completion.
Anchor text needs its own review. Replace generic wording with concise descriptions of the destination, but keep the language natural. “Shop Shopify development services” tells the user more than “find out more”, while repeating the exact same phrase in every location can make the interface awkward and over-optimised.
For a wider view of crawlability, indexability and architecture, use this guide to what a technical SEO audit covers. The internal-linking data should become an issue inventory grouped by pattern, severity and owner, not an undifferentiated list of URLs.
Prioritising Fixes by Business Value Not Just Link Count
A high link count doesn't make a page important, and a low link count doesn't automatically make it a problem. Prioritisation works better when you compare business value, current link support, crawl depth and implementation effort.
A product category with strong commercial intent but weak contextual links should normally outrank a popular blog post that already has navigation, breadcrumbs and related-content support. Conversely, adding more links to a page that already dominates the architecture can create noise and dilute attention from pages that need it.
| Page Group | Business Value | Current Inlinks and Depth | Priority Action |
|---|---|---|---|
| Core service or category pages | High | Few relevant inlinks, or deeper than three clicks | Add contextual links from relevant high-value pages, improve hubs and review navigation |
| Priority product pages | High | Underlinked compared with similar products | Link from category pages, buying guides and related products where the relationship is genuine |
| Strategic pillar guides | High | Good inlinks but weak links to supporting pages | Strengthen the cluster and create clear routes to commercial destinations |
| Supporting blog content | Medium | Many links, often concentrated in templates | Retain useful links, remove irrelevant repetition and redirect authority to priority pages |
| Documents and legacy URLs | Variable | Found through sitemap or analytics only | Confirm whether they should remain live, then link, consolidate or retire |
| Low-value utility pages | Low | Deep or lightly linked | Don't spend time adding links unless users or search engines need the page |
Use donor and destination logic
A donor page is a relevant page that can provide a useful route to another destination. A destination page is the URL you want users and crawlers to reach. The best opportunities usually connect a strong, relevant donor with an underlinked page that matters commercially.
For example, a guide about selecting accounting software may be a sensible donor for a bookkeeping service page. A general company announcement probably isn't, even if it has many internal links.
Pages at four or more clicks from the homepage deserve early review when they're strategically important. UK SEO guidance recommends bringing outreach targets within three clicks because deeper pages receive less structural support (SEO Strategy's internal linking audit checklist). That doesn't mean flattening every site into a shallow menu. It means giving priority pages a deliberate route through hubs, categories, breadcrumbs or contextual links.
Know when to stop adding links
More links aren't always better. A page overloaded with loosely related recommendations becomes harder to scan, and every additional route can weaken the information hierarchy. Keep a proposed link when it helps a user answer the next sensible question, complete a task or understand the relationship between topics.
Record the reason for each fix, its owner, implementation method and validation check. That turns subjective SEO advice into a backlog stakeholders can review and developers can execute.
Remedial Actions That Strengthen Your Site Architecture
Fixes should follow the structure you've diagnosed. Don't begin by inserting links into every article. Start with the broken routes, orphaned priority pages and deep commercial destinations, then apply patterns that can be maintained in templates.
Repair the route before adding more routes
Replace internal links to 4xx URLs with the correct live destination. If the intended page has moved, update the source link directly rather than relying on a redirect. Where several redirects form a chain, point internal links to the final canonical URL and remove outdated references from templates and content.
On an ecommerce site, a discontinued product might have links from category pages, buying guides and related products. Decide whether the product should redirect to a close replacement, remain as a useful discontinued page or be removed. Don't send every discontinued product to the homepage. That gives users a poor result and obscures the original relationship.
For duplicate or thin pages, consolidate only when the content and intent overlap. Merge the useful material into a stronger destination, update internal links, then implement the appropriate redirect. A redirect is not a substitute for deciding which page should own the topic.
Add contextual links with purpose
Content-led sites often publish a new guide and wait for it to accumulate links. A better approach is to identify existing pages that already discuss the subject and add relevant references from those pages to the new guide.
Use descriptive anchors that fit the sentence. “Internal linking audit process” is clearer than “read more”, but it shouldn't appear mechanically wherever the phrase is possible. Vary anchors according to the actual destination and surrounding copy.
For a service page, prioritise links from closely related guides, comparison pages, category pages and relevant navigation. Don't add a service link to every paragraph. A useful contextual link should make sense before the SEO rationale is considered.
Build hubs that reflect how people browse
A pillar page can introduce a subject and link to supporting articles. Those supporting pages should link back when the pillar helps the reader, creating a coherent cluster rather than a one-way directory.
On a site selling web development, a hub might organise platform, performance, ecommerce and migration resources. The hub can sit in navigation or a relevant category, while individual articles link to the hub when users need broader context.
Breadcrumbs provide another controlled route to parent categories. Related-product modules can help ecommerce users, but they need rules that prevent every product from linking to every other product. Template links multiply quickly, so test whether the recommendations are relevant and whether they are competing with more important navigation.
Watch this walkthrough for a visual explanation of internal linking principles:
Apply changes at the right level
Manual edits suit a small number of high-value pages. Template changes suit repeated structural relationships such as product-to-category or article-to-hub links. Automated systems can suggest opportunities, but automatic insertion based only on keyword matches often produces repetitive, unhelpful anchors.
Validate every implementation with a fresh crawl. Check that the new link is in the rendered HTML where appropriate, uses a valid destination, has suitable anchor text and doesn't create a loop or an unintended indexable URL pattern.
Measuring Impact and Reporting Your Internal Linking Audit
An internal linking audit is complete only when you can show what changed and whether the site now routes users and crawlers more effectively. Keep the baseline export, the amended version and a change log. Without those three items, later movement becomes difficult to interpret.
Track structural measures first:
- Crawl depth distribution: Compare how many important pages sit at each depth level.
- Unique inlinks: Monitor support for priority service, product, category and pillar pages.
- Orphan inventory: Recheck sitemap, analytics and Search Console URLs against the crawl.
- Error routes: Confirm that broken links, redirects and chains have reduced.
- Anchor quality: Review whether destination wording is descriptive and consistent.
- Organic signals: Watch indexation and visibility for the affected page groups without attributing every movement to internal links alone.
Report decisions in plain English
A useful report says what was found, why it mattered, what was changed and what remains. It can include a crawl-depth view, a list of priority pages with their link support, examples of repaired routes and a short implementation backlog.
Don't present link counts as a success metric by themselves. A rise in links can mean the architecture improved, or it can mean a template added irrelevant links everywhere. Explain the user journey and business purpose behind each major change.
For teams building automated reporting, Sibley Digital's automated SEO reports guide provides useful context on turning recurring data into a repeatable reporting process. The same discipline applies when measuring ad impact with AI agents, where the important question is not the volume of activity but whether the measurement connects work to an outcome.
Keep the link map under review
Use a cadence that matches publishing and development activity. A site with frequent content releases, migrations or catalogue changes needs more regular checks than a stable brochure site. Re-run the crawl after major template, navigation or URL changes, then review the wider structure during scheduled governance work.
The UK public sector provides a useful assurance mindset here. The Government Internal Audit Agency reported 1,336 internal audit engagements delivered to at least draft report stage in 2020–21 (GIAA annual report and accounts). Internal linking doesn't require that institutional scale, but it benefits from the same principle: define controls, record findings, assign remediation and repeat the review.
If your site has weak navigation, orphaned commercial pages or unclear crawl paths, Sibley Digital can audit the link graph, restructure the architecture and implement fixes across content and templates. Visit Sibley Digital to discuss an internal linking audit built around discoverability, accessibility and the pages that matter to the business.