Internal Links: The Architecture Google Understands (and Almost Nobody Builds)
Internal links are the only ranking factor you control 100%, for free, without asking anyone's permission — and most websites leave them to chance. Woven well, they tell Google which page matters most on each topic and route authority exactly where you need it.
May 29, 20245 min readIn this article
Of all the factors that move rankings, there's exactly one you control completely, that costs no money and depends on nobody: how you link your own pages to each other. And it is, by far, the most neglected — most websites link by inertia (the menu, the footer, the odd "see more") and let Google guess which page matters.
Internal links do three jobs at once: they guide crawling (Google discovers pages by following links), they distribute authority (every link transfers part of the emitting page's strength) and they declare hierarchy (the page receiving the most internal links is, in Google's eyes, the topic's most important one). This guide turns those three jobs into a deliberate architecture.
The pillar and cluster model: the shape of authority
The architecture that works best for businesses is simple: one pillar page covering the whole topic ("the SEO guide for small businesses") and a group of satellite articles going deep on each subtopic. Satellites link to the pillar; the pillar links to the satellites; satellites link to each other when context calls for it. The result is a topical neighbourhood where authority circulates and concentrates on the page competing for the big term.
The link text: say what's on the other side
- Descriptive, not generic: "the long tail keywords guide" tells Google (and the reader) what they'll find; "click here" says nothing.
- Natural, not robotic: the perfect anchor is the sentence you would have written anyway. Repeating the exact same keyword in every link looks like manipulation — because it is.
- Varied: if ten links point to your pricing page, let the wording vary the way real speech varies.
- Honest: a link that promises one thing and leads to another burns the reader's trust — and the behaviour signals show it.
Orphan pages and click depth
An orphan page — one no internal link points to — is invisible twice over: Google takes longer to discover it, and when it does, it reads that nobody on your own site considers it important. The first fix in any audit is giving orphans a home. The second measure is depth: how many clicks separate a page from the homepage.
Where links live (and what each spot is worth)
| Location | Example | SEO value |
|---|---|---|
| Contextual, inside the copy | A link mid-paragraph in relevant text | The highest: context + reading intent |
| Related-content modules | "Related articles" at the end of a post | Medium: useful for crawling and the next read |
| Navigation (menu, footer) | Global links across the site | Structural: defines sections, but adds no topical context |
The practical consequence: the link that counts is the one you write inside the content, where the paragraph gives it context. Menus define the map; contextual links define relevance. That's why a migration or redesign that only preserves menus loses half the architecture without anyone noticing — until the rankings drop.
A practical audit in four steps
- List your money pages — the ones that convert — and count how many internal links each receives. Your most important pages should be your most linked; if the old 2019 post gets more than your services page, there's the work.
- Hunt orphans: compare your sitemap against linked pages; every valuable page without incoming links needs at least two or three from related content.
- Review the anchors of your most repeated links: do they describe the destination, or do they say "here"?
- When publishing new content, link in both directions: from the new piece to existing ones and — the step everyone forgets — from two or three existing pieces to the new one. Without that second step, every article is born half-orphaned.
One final care: linking is concentrating, not spraying. Fifty links in one article dilute the value of all of them and get in the way of reading — the criterion is the same one that avoids duplicate content: each thing, one clear place. Done well, this architecture is among the most profitable pieces of your SEO strategy: built once, it works on every new page.
Frequently asked questions
How many internal links should an article have?
As many as context justifies — in practice, 3 to 8 in a thousand-word article is a healthy range. The criterion isn't numeric but useful: every link should answer "would the reader of this paragraph want to go deeper there?". Fifty links don't distribute authority fifty times: they pulverise it.
Do menu and footer links count?
They count for structure and crawling, but add little topical context: being on every page, they tell Google nothing specific about any of them. The fine work happens with contextual links inside content. Menu for the map, copy for the relevance.
Does heavily linking a page guarantee it ranks?
It helps, but doesn't rescue it: internal links amplify a page that already answers a search well — they don't compensate for weak content. The right order: content that deserves to rank first, architecture that channels authority to it second. The reverse is pushing a cart with no wheels.
What happens to internal links in a redesign?
It's the moment of highest risk: redesigns tend to preserve menus and lose the contextual links — and with them half the architecture. Before migrating, export the current internal-link list (any crawler produces it) and verify afterwards that the important connections are still alive. Ten minutes of comparison save months of recovery.