Internal linking: connect your pages for SEO

What is internal linking?

Internal linking refers to how links are organised between the pages of the same domain. It runs through the menu, the breadcrumb, the footer, “related articles” blocks and links placed in the text. Each internal link connects a source page to a target page, with clickable text called anchor text.

There are two families of links. Structural links are generated by the template and repeat across many pages. Contextual links are placed by hand in the body of the content. The first give the site its hierarchy, the second carry meaning.

	Structural links	Contextual links
Location	Menu, breadcrumb, footer, sidebar	Paragraphs of the content
Role	Access, depth, hierarchy	Meaning, topic proximity, next step
Risk	Dilution: the same links everywhere	Weak anchors, cannibalisation
Fix	Template redesign	Editorial linking plan

Internal linking should not be confused with link building, which refers to links received from other sites.

Why does internal linking matter to Google?

Google explores the web by following links. According to its starter guide, links connect “your users and search engines to other parts of your site” (SEO Starter Guide, December 10, 2025). A page that no link reaches, known as an orphan page, stays hard to discover, even if it is listed in the sitemap.

Internal linking also signals a hierarchy. The more links a page receives within a site, the higher its relative importance. Google says so in its documentation on ecommerce site structure (Google Search Central, updated on December 10, 2025). Linking to a page from the home page or from several pieces of content therefore marks it as a priority.

Finally, anchors and the context of links help Google understand what the target page is about. This is the basis of Google’s historical popularity calculation, covered in our article on how PageRank works.

What Google documents, and what is theory

A lot of advice on internal linking circulates without a source. Before reorganising a site, separate what Google has written from what practitioners have modelled.

Claim	Status	Source
Only <a> links with an href attribute are reliably followed	Documented	Google, crawlable links, December 10, 2025
Anchor text should be descriptive, concise and relevant; avoid “click here”	Documented	Google, crawlable links, December 10, 2025
Every important page should receive at least one link from another page	Documented	Google, crawlable links, December 10, 2025
There is no ideal number of links per page	Documented	Google, crawlable links, December 10, 2025
Silo, semantic cocoon, topical PageRank	Practitioner theory	No Google documentation
Only the first link to a page would count (“first link priority”)	Never confirmed	SEJ, July 14, 2026

The reference page is Google’s page on crawlable links and anchor text. It states that good anchor text is “descriptive, reasonably concise, and relevant”, and that no magic number of links exists. Architecture models, such as the semantic cocoon, remain working frameworks. They help you decide, without being certified mechanisms.

How do you build an internal linking plan?

An internal linking plan is designed at site level, not page by page. A link added to page B can push back page C, which answered the same query better. Hence the order of the steps below.

1. Draw up the intent → target page table: one search intent, a single page that answers it. Spot duplicates before linking anything.
2. Define the topic areas: group pages by broad subject, with a pillar page accessible from the menu.
3. Set rules per template: which links the menu, breadcrumb and automatic blocks display on each type of page.
4. Place contextual links: in the text, towards the page that goes deeper into the concept mentioned or towards the next step of the journey.
5. Write the anchors: descriptive, varied, carrying the topic of the target page.
6. Test before going live: crawl the staging site, check depth and make sure there are no broken links.
7. Roll out in stages and observe for at least 3 to 4 weeks before drawing conclusions.

The first step relies on analysing search intent. Two pages that target the same intent compete with each other; internal linking only makes the problem worse.

Choosing anchor text

Anchor text announces what readers will find when they click. “Technical audit guide” informs, “learn more” adds nothing. Vary the wording towards the same page. Avoid repeating the same exact anchor on several menu entries: Google advises against keyword stuffing in anchor text.

How many internal links per page?

Google sets no figure and notes that if the number of links looks excessive, it probably is. At Elev8 Lab, the editorial rule sets 3 to 10 contextual links per piece of content. One of them always points to the parent page. It is an internal convention, not a Google rule: it forces you to choose useful links rather than link everything.

How do you audit your internal linking?

An internal linking audit starts with a full crawl, for example with Screaming Frog. Cross-check it with Search Console and your analytics tool. Since version 24.0, the software also detects non-crawlable links built in JavaScript (release history, version 24.3 of June 29, 2026).

Checkpoint	Working benchmark	Fix
Orphan pages	0 strategic pages with no incoming link	Link from related pages, otherwise deindex
Click depth	Strategic pages within 3 clicks at most	Bring them up via menu, breadcrumb, contextual links
Link distribution	Most linked pages = most strategic pages	Reduce links to utility pages
Anchors	No generic anchor	Rewrite to describe the target
Broken or redirected links	No internal link to a 404 or a 301	Fix the hard-coded URL

To find linking opportunities at scale, use the n-gram method. It spots recurring phrases in the text that are not yet linked. Each one is then matched with an existing target page. François Dragon described it for Réacteur (“Opportunités de maillage interne à l’échelle”, in French, October 15, 2024). Each suggestion still has to be validated by hand against the intent → page table. Very deep pages also raise a crawl budget issue on large sites.

What internal linking mistakes should you avoid?

• Non-crawlable links: a clickable <span> or a plain onclick is not a link for Google.
• Generic anchors: “read the article”, “discover”, “here” say nothing about the target.
• Links to redirects: after a redesign, replace the URL with its final destination rather than leaving a chained 301 redirect.
• Automatic linking that nobody reviews: plugins that place links on keywords repeat the same exact anchor and ignore intent.
• Internal nofollow: it does not redirect popularity elsewhere, it loses it.
• Link obfuscation: turning a link into a button to “steer” Google complicates the code for an uncertain effect.

On that last point, an SEO wanted to hide a link to favour another anchor. John Mueller (Google) replied that he suspected they were overthinking it (Search Engine Journal, July 14, 2026). He recommends keeping a proper <a href> link and handling the presentation in CSS.

A concrete example: Elev8 Lab’s architecture

The site you are reading applies these principles. Its structure has three levels. Pillars sit at the root (/seo/), sub-pillars below them, and articles are attached to their pillar. All content can be reached in three clicks from the home page.

• This article belongs to the website structure sub-pillar, linked from the SEO pillar.
• Each article links to its parent page, then to 2 to 9 pieces of content that go deeper into a specific point.
• The breadcrumb is native HTML, marked up as BreadcrumbList following Google’s breadcrumb documentation.
• Article URLs do not depend on sub-pillars: reorganising a section breaks no links.

This choice keeps structural links to the bare minimum and lets contextual internal linking carry the meaning. It will be measured in Search Console, by page group, after publication.