01Notes · 22 August 2026
Internal linking is the cheapest SEO work you are not doing
Quick answer
Internal linking needs no outreach, no approvals, no supplier and nobody’s permission. It is the only structural SEO work that is entirely inside your own control.
It is also almost universally neglected, and the reason is commercial rather than technical: nobody sells it. There is no invoice line, no dashboard and no monthly report that makes it look like progress.
Key takeaways
- Nobody needs to approve it. No outreach, no budget, no waiting on a publisher.
- Nobody sells it either, which is most of why it goes undone.
- Start with your strongest pages. A page with authority linking to nothing useful is the most common waste we find.
- Orphan pages are invisible. If nothing on your site links to a page, you have told search engines it does not matter.
- No exact-match anchor quotas. Link count showed almost no independent effect on AI citation position, and quotas are a detectable pattern.
- An hour of work usually finds something. This is the highest return-per-hour task in most audits.
The reason it goes undone
Not because it is difficult. Because there is no market for it.
Link building has suppliers, prices and a report showing what arrived this month. Content has writers and a publishing calendar. Technical SEO has tools that produce a score that goes up. Internal linking has none of that. It is a person reading their own site carefully and making decisions.
So it sits in the gap between everyone’s remit. The content team writes the page. The SEO team audits the page. Nobody owns the question of what that page should link to and what should link to it.
Topic clusters, without the jargon
The idea is simpler than the terminology suggests. One page covers a subject broadly. Other pages cover parts of it in depth. They link to each other in both directions.
That structure does two useful things. It tells search engines which page is the authority on the subject, and it stops your own pages competing with each other, which is the problem consolidation solves from the other direction.
| Element | Working | Common failure |
|---|---|---|
| Pillar page | Covers the subject broadly, links to every child | Exists, links to nothing, or was never designated |
| Supporting pages | Each covers one part in depth, links back to the pillar | Link to the homepage instead, which helps nothing |
| Sibling links | Where the relationship genuinely helps a reader | Every page links to every other page. Meaningless |
| Anchor text | Describes the destination, varies naturally | The same exact-match phrase every time |
| Placement | In the sentence that was already about that subject | A "related links" block nobody reads |
| Direction | Strong pages passing signals to pages that need them | Weak pages linking to the homepage, which needs nothing |
The last row is the one worth acting on first. Most sites have their authority pointing at pages that already have plenty.
A one-hour audit
- Crawl your own site. Any free crawler will do. Export the internal link report.
- Sort by inbound internal links. Pages with zero are orphans — you have told search engines they do not matter. Pages with hundreds are usually navigation rather than editorial signal.
- List your strongest pages by traffic or by referring domains. Then read what they link to. If a page with real authority links only to your homepage and your contact page, that authority is going nowhere useful.
- Find the commercial pages with weakest inbound support. Those are the destinations. Link to them from the strong pages, in sentences already about that subject.
- Check for redirect chains in internal links. Internal links pointing at URLs that redirect are a small, free loss repeated across a site.
An hour, no budget, no approvals. In our audits this consistently produces more actionable findings per hour than anything except crawler policy.
Anchor text, and the quota trap
Someone will eventually suggest a target: 30% exact-match anchors, or two internal links per 500 words. Decline.
Quotas produce patterns, and patterns are exactly what spam detection is built to find. In 2026 research, link count showed almost no independent effect on AI citation position — there is no number to hit.
The test that works is simpler. Would you have written that link if you were explaining the subject to someone? If yes, keep it. If you added it to hit a number, remove it.
Where this connects to AI citations
Indirectly, and worth stating precisely rather than overclaiming.
Internal linking does not cause AI citations. What it does is help search engines identify which of your pages is the authority on a subject, which supports ranking, and ranking remains the strongest single gateway to citation. Pages holding traditional top-three positions appeared in AI citations 98.9% of the time.
So the chain is: better internal structure, clearer authority signals, better ranking, higher citation likelihood. Three steps, each real, none of them a shortcut.
What this means for you
- Do the one-hour audit before your next content sprint. It costs nothing and it usually redirects the sprint.
- Start with your strongest pages, not your weakest. Authority pointing nowhere is the most common waste.
- Refuse any anchor-text quota. There is no number, and quotas create the pattern you want to avoid.
02Common questions
Questions this raises
Why is internal linking undervalued?
Because it is unglamorous and nobody sells it. There is no supplier, no invoice line and no dashboard. It is also the only off-page-shaped work that requires nobody's permission but your own.
What is a topic cluster?
A pillar page covering a subject broadly, with supporting pages covering parts of it in depth, linked in both directions. The structure tells search engines which page is the authority on the subject and which are supporting.
How many internal links should a page have?
There is no correct number and anyone quoting one is guessing. The useful question is whether every link earns its place by helping a reader. Link count showed almost no independent effect on AI citation position in 2026 research.
Should anchor text be exact-match?
No. Vary it naturally. An exact-match anchor quota is a pattern, and patterns are what spam systems are built to find. Describe the destination the way you would in conversation.
Does internal linking help with AI citations?
Indirectly. It helps search engines understand which of your pages is the authority on a topic, and traditional ranking remains the strongest gateway to AI citation — top-three pages appeared in AI citations 98.9% of the time.
What is the fastest internal linking win?
Find your strongest pages by traffic or links, and check what they link to. Strong pages passing signals to nothing useful is the most common waste we find, and it takes an hour to fix.
How do I audit internal links without a tool?
Crawl your own site with any free crawler, export the internal link report, and sort by inbound links. Pages with zero internal links pointing at them are orphans. Pages with hundreds are probably navigation, not editorial.
Sources
Every figure on this page comes from one of the following. Where a claim is ours rather than a source’s, it carries an evidence label instead.
- Rankability — State of AI Search, 1,645 citation observations across 31 topics, 2026. Source for the 98.9% top-three figure and the finding that link count showed almost no independent effect on citation position. Independent study
- Google Search Central — spam policies covering link schemes and detectable patterns, now explicitly in scope for generative AI responses. Primary
- Searlixa delivery across client engagements since 2018, labelled Observed with sample and date. Our own
Ask an assistant
Hand this page to an assistant and ask it to summarise or challenge the argument. If it gets something wrong about us, that is itself a useful finding.
→Next step