Internal linking that actually moves rankings

Internal links do three jobs at once. They let crawlers discover pages, they tell search engines which pages you consider important, and their anchor text and surrounding words explain what each target page is about. Unlike backlinks, you control every one of them. Yet on most sites they are an accident of the CMS: whatever the theme puts in the menu, whatever the "related posts" widget picks, and whatever a writer remembered to add.

This guide is about turning that accident into a structure. It assumes you already have reasonable content; links cannot rescue pages that do not deserve to rank, but they very often hold back pages that do.

How internal links influence rankings

Search engines treat a link from page A to page B as a vote that B matters, weighted by how much A itself is linked to. The home page usually collects the most external links, so pages one click from the home page inherit more than pages five clicks deep. That is the authority side.

The relevance side comes from the anchor text and the text around the link. A link reading "our guide to choosing a trail running shoe" inside a paragraph about ankle support tells a search engine far more than a "click here" in a footer. Google has said for years that anchor text helps it understand the target page, and that applies to internal anchors too.

There is also a crawling side. Pages that are hard to reach get crawled less often, so updates to them are noticed later. For large sites that interacts with crawl budget; for small sites it mostly means new pages sit unindexed for longer than necessary.

Measure before you change anything

Run a crawl of your own site with any desktop or cloud crawler and export three things for every indexable URL: click depth from the home page, the number of unique internal pages linking to it (inlinks), and the anchor texts used. Then join that with clicks and impressions from Search Console. You now have a table you can sort.

Pattern in the data	What it usually means	Typical fix
High impressions, average position 8–20, few inlinks	Google already sees relevance but the page lacks internal support	Add 3–10 contextual links from related, well-linked pages
Click depth 5 or more for important pages	Pagination or deep category trees bury them	Link from hub pages, shorten category paths, add "popular" blocks
Zero inlinks, present in sitemap	Orphan page	Link it from a relevant page, or decide it should not exist
Hundreds of inlinks to a low-value page	Sitewide template link (login, cart, a tag)	Usually fine; just do not mistake it for importance
Same anchor text pointing to two different pages	Mixed signals, often cannibalisation	Pick one target per topic; see keyword cannibalization

Orphans deserve a separate pass. Compare the URLs in your sitemap and analytics with the URLs your crawler found by following links. Anything in the first list but not the second is either orphaned or reachable only through JavaScript the crawler did not execute; the JavaScript guide explains how to tell the difference.

Navigation links versus contextual links

Both matter, but they do different work.

• Navigation (main menu, breadcrumbs, footer, category sidebars) defines the site's skeleton. It should make every important section reachable within two or three clicks and stay stable over time. Mega-menus that link to 300 pages from every page flatten the structure so much that nothing stands out.
• Breadcrumbs are underrated. They give every page a link back up its hierarchy with clear anchor text, and they are a cheap way to make deep pages less isolated.
• Contextual links inside body copy carry the strongest relevance signals. A link placed where a reader would naturally want more detail on a subtopic is the ideal internal link.
• Automated blocks such as "related articles" help with discovery but are often poor at relevance, because they match on tags or recency. Check what yours actually picks for a few important pages.

Anchor text that helps

A good internal anchor describes the target in the words a searcher would use, reads naturally in the sentence, and differs a little each time. For a hypothetical outdoor-gear store linking to its buying guide for down jackets, these work well: "how fill power affects warmth", "our down jacket buying guide", "choosing a down jacket for wet climates". These do not: "click here", "read more", "this article", or the exact phrase "best down jacket 2026 buying guide" repeated on forty pages.

Two practical rules keep anchors honest. First, write the sentence for the reader and then see whether the natural phrase is already a good anchor; it usually is. Second, never link the same anchor text to two different pages, because you are telling search engines that both pages are the answer to the same thing.

Image links use the image's alt attribute as anchor text. A logo linking to the home page with empty alt text is an anchorless link; give it the brand name.

A routine for linking new content

Most internal linking decays because new pages are added without anyone going back to older pages. A five-minute routine for every new piece prevents that.

1. Link out from the new page to two to five existing pages a reader would need next: the parent topic, a deeper guide on a subtopic, the relevant product or tool.
2. Find pages that should link in. Search your own site for the main topic phrase (a site: search or your CMS search). Pick the three to five most relevant, well-trafficked pages.
3. Add a contextual link on each, in a sentence that already discusses the subtopic. If no such sentence exists, add one that genuinely helps the reader.
4. Check the hub. If the topic has a hub or category page, make sure the new page is listed there.
5. Record it. A simple spreadsheet of "new page, pages linking in" makes later audits much faster.

Common mistakes

• Linking to redirected URLs. After a migration, menus and old articles often still point to old addresses. Each link costs an extra hop and dilutes signals. Update them to the final URL.
• Nofollowing internal links to "sculpt" authority. It does not conserve anything; the authority assigned to a nofollowed link is simply lost. Remove links you do not want instead.
• Hiding links in tabs or accordions that load on click via JavaScript. If the link only exists in the DOM after an interaction, some crawlers never see it. Content collapsed with CSS is fine; content fetched on click is not.
• Paginated archives as the only route to older posts. Page 14 of a blog archive is a long way from anywhere. Topic hubs and contextual links shorten the path; see pagination and infinite scroll.
• Linking everything to everything. A page with 400 contextual links signals nothing about which ones matter. Fewer, better-placed links are worth more.

How to tell whether it worked

Pick ten pages that were in positions 8 to 20 with few inlinks, add contextual links from relevant pages, and leave a control group of ten similar pages untouched. Check average position and clicks for both groups after four to six weeks, using the same date windows. If the test group moves and the control group does not, roll the same treatment out to the next batch of pages. Internal link changes are among the faster-acting SEO changes because they only require a recrawl of the linking pages, but results vary, and a control group protects you from crediting a seasonal swing or an algorithm update.

Keep the structural basics in check while you do this. A quick run of the free SEO checker on a few templates will flag broken internal links, redirect chains and pages blocked from indexing, which no amount of careful anchor text can compensate for.