Internal linking is how pages on your own site link to each other, and it is the ranking lever most sites never pull deliberately. Backlinks must be earned from strangers; internal links you place yourself, today, and they decide three things Google cares about: which pages get crawled, where your site's authority pools, and what each page appears to be about. Most sites let all three happen by accident. This post covers how link equity actually flows, where on the page a link matters most, the anchor rules worth enforcing, the hub and spoke architecture we build client sites on with a concrete example, and a six-step audit that usually surfaces rankings a site already paid for.
The short version
- Internal links steer crawling, authority flow, and topical meaning, and all three are fully in your control.
- Equity divides across the links on a page and flows strongest through contextual body links.
- Anchors describe the destination; one authored link per target URL per page.
- Structure follows hub and spoke, where every page supports exactly one hub.
- Never nofollow an internal link, and aim new links at pages ranking 5 to 15.
How authority actually moves through your site
Google discovers pages by following links, so your internal links are the map it crawls. The same paths carry authority. In the classic PageRank model, still the most useful mental picture anyone outside Google has, a page's equity divides across its outgoing links: a link from a page carrying forty links hands over a meaningful share, while a link buried among four hundred passes a sliver. Your homepage is almost always the strongest page you own because most external links point there, which means every link it carries is prime real estate, and equity thins with each hop away from it. Anchor text rides along as a label: when five pages link to one URL with anchors about local SEO, Google needs no further hint about that page's topic. Google's search advocates have called internal linking one of the most powerful things a site owner controls, which is unusual directness from a company that rarely ranks tactics for you. The reason it goes unpulled is that link building gets the budget while the equity a site already earned pools in the wrong places.
Where a link lives changes what it does
| Link location | What it does well | Its limit |
|---|---|---|
| Main navigation | Defines the site's structure and keeps core pages one click from everything | Sitewide repetition dilutes it as a topical signal; reserve slots for pages that earn them |
| Breadcrumbs | Encode hierarchy for users and crawlers, feed BreadcrumbList schema | Fixed anchors, no editorial signal beyond the structure itself |
| Contextual body links | The strongest signal you control: chosen anchor, surrounding relevance, editorial intent | Requires actual writing; nobody automates it well |
| Footer links | Utility access to legal, contact, and hub pages | Heavily discounted as a ranking signal; stuffing keywords here is an old spam pattern |
| Related-content modules | Scale crawl paths across large sites automatically | Algorithmic picks carry weaker intent than a link a human placed in copy |
The working rule falls out of the table: navigation and breadcrumbs set the skeleton, but the links that move rankings are the ones placed by hand in body copy, where the anchor and the surrounding paragraph both vouch for the destination.
Anchor discipline
The anchor is the link's job description. These are the rules we enforce in every piece of copy we publish.
- "Click here" and "read more" tell crawlers and screen readers nothing. The destination's topic vanishes from the signal.
- Naked URLs as anchors waste the label entirely; a URL is an address, not a description.
- Linking the same page five times in one article adds nothing after the first link. Long-standing crawler behavior credits the first anchor it meets, so repeats are clutter with no signal.
- Descriptive anchors that name the destination's topic, written so the sentence reads naturally aloud.
- Exact-match anchors on your own site. The manipulation caution that applies to backlink anchors does not transfer; your site describing its own pages accurately is what accuracy looks like.
- The same discipline everywhere you write. The anchor rules travel with you into the articles you place on other sites; our post on guest blogging covers that side.
The hard rule in the Egochi style system: one authored link per target URL per page. Once a URL is linked in body copy, every later mention renders as plain text. Navigation, breadcrumbs, and repeated CTA buttons are interface, not copy, and sit outside the rule. Our build fails if a page breaks it. The payoff is that every link on the page is a deliberate signal instead of habit, and no page leaks equity through a dozen redundant repeats.
Hub and spoke is the structure that compounds
Scattered links help pages; architecture builds authority. The structure that compounds is hub and spoke: one hub page owns the broad commercial topic, and a cluster of spoke pages each owns one specific keyword underneath it. Spokes link up to their hub with descriptive anchors, the hub links down to every spoke, and spokes cross-link where a reader genuinely benefits. Search engines read that shape as depth: one page claiming a topic is an assertion, a linked cluster covering it is evidence.
Made concrete for an HVAC company: /services/ is the hub that targets HVAC services. Under it, /ac-repair/, /furnace-repair/, and /duct-cleaning/ each own exactly one service keyword and link up to the hub. Blog posts like "why is my AC blowing warm air" each support one spoke and link to it with an anchor naming the service, never generically to the homepage. When a prospect or a crawler lands anywhere in that cluster, every path leads through the topic, and the hub accumulates the combined weight of everything beneath it. This is not theory to us. Egochi's own site runs on the same discipline, every page targets one keyword and links up into exactly one hub, a rule that also kills keyword cannibalization before it starts, because two pages fighting over one query is an architecture bug, not a content problem. The map scales in one direction only: a new page joins an existing cluster or starts a new one with its own hub, and no page ever floats free. We apply the structure in our SEO services engagements, and it is usually the change that moves mid-ranking pages the most within the first quarter.
Click depth, orphans, and the nofollow mistake
Three structural failures undo otherwise good sites. Deep pages: anything more than three clicks from the homepage gets crawled less and read as less important, whatever its quality. Orphans: pages with no internal links at all, invisible to crawlers following your site and cut off from authority flow. And nofollowed internal links, usually left over from an old PageRank sculpting experiment: since Google changed how the attribute handles equity, a nofollowed link evaporates the authority it would have carried instead of redistributing it, so sculpting throws value away. Our post on nofollow links tells that story in full; the short rule is that internal links are always normal links. All three failures are silent, which is why they persist for years, and all three show up in minutes with a crawl.
The six-step audit
-
Crawl the site
Run Screaming Frog or a comparable crawler and export every internal link with its anchor text. This inventory is the raw material for everything below.
-
List the orphans
Compare crawled URLs against your sitemap. Pages in the sitemap that no internal link reaches are orphans; link each from a relevant page, merge it, or retire it.
-
Fix click depth
Flag pages deeper than three clicks. Pull the ones that earn money or traffic up through hub links or navigation.
-
Clean the anchors
Search the export for "click here", "read more", "learn more", and naked URLs. Rewrite each to name the destination's topic, and strip any rel="nofollow" you find on internal links while you are in there.
-
Count in-links to money pages
Sort pages by internal links received. If a blog post has forty and your highest-margin service page has six, your authority is pooling in the wrong place. Rebalance toward the pages that pay you.
-
Point links at strikers
Pull queries ranking 5 to 15 from Search Console and add descriptive-anchor links to those pages from your strongest related content. This is where internal linking pays back fastest, because the target already almost ranks.
Where this fits in on-page work
Internal linking is one pillar of page-level SEO, next to titles, headings, content depth, and the markup that makes a page legible to machines. Our on-page SEO services cover the full set, and the internal-link pass is where most engagements find the quickest wins, because every fix ships the day you decide it.
Questions people ask about internal linking
How many internal links should a page have?
As many as genuinely help the reader, which for most pages lands between 3 and 10 in the body copy. There is no magic count. The rule that matters more is one link per target URL per page: once you have linked a page from the copy, later mentions render as plain text.
Do internal links help SEO directly?
Yes, three ways. They create the crawl paths Google follows to find pages, they pass authority from strong pages to pages that need it, and their anchor text tells Google what the target page is about. Google's own advocates have called internal linking one of the most powerful things a site owner controls.
Can internal anchor text be exact match?
Yes, and this is where backlink caution misleads people. Over-optimized anchors from other sites can look manipulative; your own site describing its own pages accurately cannot. "SEO pricing" pointing at your pricing page is exactly what both readers and crawlers want. Vary the wording for readability, not for fear.
Should I nofollow any internal links?
No. Nofollowing internal links throws authority away instead of steering it, because Google treats the equity a nofollowed link would have carried as evaporated rather than redistributed. Pages you truly do not want indexed need a noindex tag, and pages you do want indexed deserve normal links.
What is an orphan page?
A page with no internal links pointing to it, reachable only through the sitemap or from outside. Crawlers find orphans late or never, and no authority flows to them, so they rank far below their potential. Every audit should end with a decision for each orphan: link it, merge it, or remove it.
How fast do internal linking changes work?
Faster than almost any other SEO change, because everything is under your control. Once Google recrawls the linking pages, typically days to a few weeks, the signals update. Movement shows first on queries where the target already ranks between positions 5 and 15, which is exactly where to aim.
