Internal Links Are a PageRank Distribution System. Treat Them Like One.
Most SEO audits treat internal linking as a content discovery problem. Is the page reachable? Does it have links? Move on. That framing misses the actual mechanism — and the actual cost of getting it wrong.
Google’s link graph assigns PageRank to every page. Internal links don’t just signal relevance; they split and transfer equity. Every outgoing link on a page divides the PageRank that page holds and passes a fraction to each destination. That’s the original PageRank formula from the 1998 Brin and Page paper, and despite 25 years of updates, the core math still applies. The problem is that most sites have quietly built internal link structures that haemorrhage this equity in ways that never show up on a standard site audit.
Where the Dilution Actually Happens
Take a large e-commerce or content site: a global navigation linking to 80 category pages, a footer with 40 links, a sidebar with 20 related posts, breadcrumbs. A single page might carry 150+ outgoing internal links before you’ve added one contextual link in the body copy. Every one of those links takes a slice of whatever PageRank that page holds and routes it somewhere.
The math is blunt. If a page has 150 outgoing links and holds 1.0 units of distributable PageRank (after the classical 0.85 damping factor), each destination gets roughly 0.0057 units from that page. Add five contextual in-body links and those five real destinations are now competing with 145 boilerplate links for the same pool of equity.
Navigation and footer links are the biggest culprits. They appear on every page, which means they receive internal PageRank from your entire site — but they also dilute every page they appear on simultaneously. If your global nav has 80 links across 5,000 pages, those 80 destinations accumulate enormous internal equity not because you deliberately sent it there, but because a developer thought a mega-menu was user-friendly.
The Boilerplate vs. Contextual Link Distinction
Google’s Reasonable Surfer patent distinguishes between links likely to be clicked and links that exist as structural boilerplate. Prominent in-body contextual links in actual content pass more signal than footer links buried in a grid of 40 destinations. That’s not just about click probability; it’s about how Google weights both the equity transfer and the relevance signal attached to the anchor text.
So the problem is double: you’re diluting PageRank with volume, and the links that actually matter aren’t getting full relevance credit because they’re drowning in noise.
How to Audit This (Not Just Count Links)
A standard crawl report tells you a page has 150 internal links. It doesn’t tell you how much equity is flowing where, or whether your most important pages are actually receiving enough of it. You need two things: link topology data and PageRank simulation.
Step 1: Map Your Link Topology by Link Type
Pull a full crawl with Screaming Frog or a custom crawler. Segment internal links by position: navigation, footer, breadcrumb, sidebar, body copy. Screaming Frog handles this via the Custom Extraction tab with XPath targeting specific DOM containers.
Once segmented, count outgoing links per page by type. Any page with more than 100 total outgoing internal links is almost certainly a dilution problem — especially if fewer than 20% of those are in-body contextual links.
Step 2: Run a PageRank Simulation
Tools like Sitebulb approximate internal PageRank distribution. It’s not the real thing — Google’s actual implementation includes damping, Reasonable Surfer weighting, and factors we don’t have access to — but the approximation is useful for comparative analysis. You’re looking for pages that rank well for competitive terms but show very low simulated PageRank. That gap is often link dilution, not content weakness.
We’ve used this approach to find core commercial pages receiving roughly the same simulated internal equity as tag archive pages — because the nav linked to both at similar frequency, and the tag page had been pulled into the link graph via faceted navigation. The fix wasn’t adding more links to the commercial page; it was removing unnecessary destinations competing for the same equity pool.
Step 3: Find Equity Sinks
An equity sink is a page that receives a lot of internal PageRank but passes very little forward — usually because it has no outgoing links, links to noindexed pages, or links to pages that are themselves dead ends. Paginated archives with no contextual links, tag pages, author pages: common culprits. These absorb equity from your site and contribute nothing back to the flow.
The Fix Is Not “Add More Internal Links to Important Pages”
This is the mistake most practitioners make when they recognize the problem. They add 10 internal links to a target money page from various posts and call it done. Sometimes that works. But if every page you’re linking from still has 150 outgoing links, you’re adding a small amount of equity to a still-leaky system.
The real fix is structural. Reduce boilerplate link volume. A global nav with 80 links should be closer to 20. Footer links should be deliberately chosen, not comprehensive. Sidebar “related posts” widgets that auto-populate with 15 items should probably show 5, chosen by relevance, not recency.
Consolidate where you can. If you have 12 tag archive pages for minor topic variations, those pages are all accumulating internal equity and producing nothing useful. Noindex them, remove them from the internal link graph entirely, and that equity flows instead to pages that deserve it.
One Caveat Worth Saying Out Loud
Google’s actual PageRank implementation is not public. The 1998 paper is a foundation, not a specification. The Reasonable Surfer patent, Topic-Sensitive PageRank, and subsequent system updates all modify how link equity flows in ways we can only approximate. If someone sells you a precise number for how much PageRank a specific internal link passes, they’re fabricating it.
What we can reason about reliably: pages with fewer outgoing links pass more equity per link, contextual in-body links carry more weight than boilerplate links, and equity sinks are a real drag on large crawlable sites. The mechanism is real even if the exact coefficients aren’t knowable.
The Structural Question to Sit With
Count the unique destinations in your site’s navigation and footer right now. If the number is over 50, ask yourself which of those you’d deliberately choose to give internal PageRank to if you had to pick manually. Anything that wouldn’t make that list probably shouldn’t be in your global link template — because right now, it’s on every single page you publish, taking a cut of every page’s equity, and you never decided that was where it should go.