Discovery happens through links
Google says the vast majority of new pages it finds each day come from links. A page nothing links to is relying on a sitemap alone, which is a far weaker introduction — and a sitemap entry, unlike a link, carries no context about what the page is for.
That difference is the reason internal linking is not an optimisation applied afterwards. It is the mechanism by which a site is discovered at all, and the one part of link acquisition that needs nobody's permission. You cannot decide who links to you from outside; you decide entirely how your own pages connect.
Anchor text is a description
Link text tells readers and engines what to expect on the other side. "Read more" describes nothing, and "click here" describes less. A short phrase naming the destination's topic describes it to both audiences at once, and costs no extra words.
The instinct to over-optimise here is worth resisting. Anchor text that repeats the same exact phrase on every link into a page reads as engineered rather than written, and the natural version — varied phrasing that still names the topic — is both better prose and a better signal. Write the link the way you would say it aloud.
| Anchor | What it communicates |
|---|---|
| Read more | Nothing about the destination |
| Click here | Nothing, plus a device assumption |
| this guide | That something is a guide, and no more |
| how hreflang return links work | The topic, the depth, and why to click |
Depth and direction
Google's practical requirement is simpler than a click-depth formula: every important page should have at least one link from another page that Google can crawl. Pages that are difficult to reach are also easier to orphan accidentally and harder for users to discover.
Link forward from hubs to useful detail pages and back where that helps navigation. Contextual links are valuable because descriptive anchor text and surrounding copy explain why the destination matters; site-wide navigation can still be useful when it genuinely helps people move around the site.
- List the pages that actually matter — the ones worth traffic.
- For each, find how many clicks separate it from the home page.
- Where that number is large, ask which existing page should be linking to it.
- Add the link in body content, with anchor text naming the topic.
- Check the reverse direction exists too, so the pair is connected both ways.
Where internal linking quietly goes wrong
Most sites do not have a linking strategy problem. They have a small number of specific defects that accumulated without anyone deciding on them.
| Defect | Effect |
|---|---|
| Links only in the main navigation | Every page gets the same generic treatment |
| Internal links pointing through redirects | Wasted crawl and a diluted signal |
| Links to non-canonical variants | Signals split across URLs that should be one |
| nofollow on your own pages | Withholding a vouch from pages you control |
| Pagination that hides deep items | Items past page two effectively unreachable |
Linking as a description of your site
The shape of your internal links is an argument about what your site is about. A page with twenty links pointing at it from relevant contexts is being described, repeatedly, as important to that subject. A page with one link from a footer is being described as an afterthought, because that is what it is.
This is also why linking works best when it follows the content rather than preceding it. Deciding "this page needs fifteen internal links" produces fifteen contrived sentences. Writing the page properly and linking wherever a reader would genuinely want the other page produces fewer links that each mean something — and the count was never the target.