Internal links

Internal links are links from one page of your site to another page of the same site. They are the least glamorous thing in search, entirely within your control, and responsible for more quiet underperformance than almost anything else. A page nothing links to can still be found through your sitemap, and it arrives with nothing on the site suggesting it matters.

In short

• They are links between your own pages, and you control every one of them.
• They are how new pages get discovered and how importance is signalled.
• A page nothing links to is an orphan, and it usually performs badly.
• Links written into sentences count for more than menu items.

The three jobs they do

They do three jobs, and most advice about them only mentions the first.

• Discovery. Following links is how search engines find pages fastest. A sitemap will get a page noticed eventually, and a link gets it noticed now.
• Description. The wording of the link tells search engines what the destination is about, which is the subject of anchor text.
• Distribution. Whatever standing a page has is passed along its links, so the pages you link to most from your strongest pages tend to do better.

The third is where the folklore lives. It is real, it is not a lever you can pull precisely, and treating it as one produces sites engineered for a formula and unpleasant to read.

Why the menu is not enough

Almost every business site links to its service pages from the navigation and nowhere else, and treats the job as done.

The trouble is that a menu appears on every page identically. It tells a search engine that these pages exist and that they matter to the site in general. It says nothing about which page is relevant to what, because it is the same on the contact page as on the article about the subject.

A link written into a sentence is different. It appears on one page, in a context about that subject, with wording chosen for that moment. That specificity is what gives it more weight, and it is the linking most sites have none of.

What a link actually passes

A page that other pages point at is treated as more important than one nothing points at, which is the whole idea underneath this and is routinely described with more mystique than it deserves. That importance flows along links, so a page linked from your home page inherits more than a page linked from an archive on page four. It is the same idea as a link from another site, running inside your own.

Two things follow. Pages you want to rank should be linked from pages that are already doing well, rather than from wherever was convenient. And a page linking out to two hundred destinations passes very little to any of them, because whatever it had is divided among them.

Orphans, and how sites get them

An orphan is a page nothing on the site links to, and they accumulate in predictable ways rather than through carelessness. A page gets built for a campaign that ended. A service is added and the menu was never updated. A blog post falls off the third page of an archive nobody reads. A page is renamed, and the old links point somewhere else now. None of these feel like mistakes at the time, and all of them produce a page the site has quietly disowned.

The fix is a list. Compare the pages in your sitemap against the pages anything actually links to, and the difference is your orphan list. Most owners are surprised, and usually at least one of the orphans is a page they care about.

Linking backwards is the step people skip

When a new page goes up, linking out from it is instinctive. Linking to it is the part that gets forgotten, and it is the half that matters.

The habit worth building is short. Before publishing, find the two or three existing pages where the new subject genuinely comes up, and add a link from each, worded naturally. That is what tells a search engine the page exists, what it is about, and that the rest of the site considers it relevant.

Doing it later is harder, because by then nobody remembers which pages those were. Doing it never is how a perfectly good page sits at position forty for two years.

The habits that keep it healthy

Internal linking fails through neglect rather than difficulty, so the useful habits are the ones somebody will actually keep.

• Put them in the sentence that raises the subject, where a reader is already thinking about it, rather than in a strip at the foot of the page.
• Describe the destination in the wording, since that is what the link is telling a search engine.
• Send links towards the pages that earn money, rather than spreading them evenly for tidiness.
• Check for broken ones occasionally. Renamed pages leave links pointing at nothing, and nobody notices from the front end.
• Revisit after a redesign. Rebuilds routinely drop body links while keeping the navigation, which quietly undoes years of this work.

Where structure and linking meet

Internal linking is only the mechanism, and what to link is a different question answered well before any linking happens.

Deciding which pages should exist, what each one is for, and how they group together is information architecture. Arranging a set of pages around one subject so they support rather than compete is a topic cluster. Both produce the map. Internal links are how the map gets drawn on the site itself, and a good structure with no links between its pages is a plan nobody executed.

That order is worth respecting. Adding links to a site whose pages overlap and compete will link the confusion together rather than resolve it.