Internal linking

Internal links do three separate jobs, and they are worth separating because a site can be fine at one and poor at another. They are how a crawler finds pages in the first place. They are how importance moves around a site. And they are how a reader gets from one relevant thing to the next.

Discovery is the blunt one. A page with no internal links into it can only be found through a sitemap or an external link, and a crawler that reaches it that way has been given no context about what it relates to. Content management systems produce these routinely: the publish step creates the URL and links it from nowhere.

Importance is the part that resembles the external link story. Standing that arrives at a homepage does not stay there; it moves through the links out of it. A commercial page three clicks down with one link into it receives very little of it, and no amount of work on that page changes what the structure is saying about it.

The third job is the one clients feel. A reader who finishes an article and has nowhere obvious to go leaves, and the next thing they read is on somebody else site.

Navigation links are on every page, which is exactly what limits them. A link present a thousand times says nothing about any particular page, and the anchor is a menu label chosen for brevity rather than description.

A contextual link carries three things a navigation link does not: descriptive anchor text, the subject matter of the paragraph around it, and the implication that somebody writing about this topic thought that page was worth pointing at.

This is why an audit that counts internal links is close to useless. A page with two hundred links into it from a footer and none from body copy is in worse shape than a page with six contextual ones, and the count says the opposite.

The practical read for agency work: when a client asks for a link to a service page, the answer is not to add it to the menu. It is to find the three articles where somebody writing about the subject would naturally have mentioned it.

A glossary is the cleanest internal linking structure available to most agencies, and it is one of the few that is worth doing for reasons that have nothing to do with search.

The mechanics are simple. Articles use terms; each term has a page defining it; the article links the term to the definition with the term as the anchor. The reader who did not know the word gets an answer without leaving the site to find one. The definition page gets links from every article that uses it, with a perfectly descriptive anchor, from pages that are genuinely about the topic.

It also scales by itself. Every new article adds links to existing definitions, and every new definition gives the existing articles somewhere relevant to point. Nothing has to be maintained as a separate project.

The failure mode is mechanical application: auto-linking the first occurrence of every term on every page produces a wall of blue and reads as automation. Linking a term where the reader might actually need it, once, is the version that works.

Internal linking is almost always found in a quarterly audit, which is the most expensive time to find it. Three hundred published articles needing two links each is a week of somebody time, produces no new content, and is very hard to sell as a line on an invoice.

Two links at the moment of writing cost nothing, because the writer already knows what the article relates to. Putting it in the publishing checklist, next to the meta description and the image alt text, converts a project into a habit.

Two questions do most of the work. What existing page should this one point at, and what existing pages should now point at this one. The second is the one people forget, and it is the one that stops orphans being created.

For a client site, this belongs in whatever brief the content is written to. An agency that fixes internal links quarterly is being paid to clean up a process it could have specified once.