Internal linking is the most leverage-able SEO lever that costs nothing to change. Every page you already publish either passes authority forward or it doesn't; every page either accepts authority from hubs or it doesn't. Professional tools measure this — Screaming Frog, Ahrefs, Sitebulb — and they cost real money.
The Internal Link Graph is a free, browser-native version that does the one thing that matters for most operators: crawl a site from a seed URL, chart the internal topology, and surface orphans and hubs. No account. No install. You give it a URL, pick a page limit, click crawl, and watch.
What Counts as an Orphan, a Hub, and a Dead End
Three categories produce the highest-impact fixes and the tool calls them out explicitly:
Orphans are pages with zero inbound internal links from other crawled pages. They were reachable from your seed (otherwise the crawler wouldn't have seen them) but nothing points to them. This usually means the page is in your sitemap but nothing in the site's navigation or body copy links to it. Orphans earn close to zero internal authority. They also tell Google the page is less important than the pages that are linked.
Hubs are pages with 3 or more inbound internal links. These are where authority concentrates. Your homepage is usually the biggest hub. A well-structured site has a handful of secondary hubs — category pages, cornerstone articles — that distribute authority outward to leaf pages. If your site has only one hub (the homepage), every leaf page sits exactly two hops from the front door, which is fine for small sites and limiting for larger ones.
Dead ends are pages with zero outbound internal links. A blog post that doesn't link to any related post is a dead end. Crawlers stop there and so do readers. Dead ends are not a ranking penalty, but they waste the authority a hub passed to them — it has nowhere to flow.
How the Crawl Works
You pick a seed URL (usually the homepage), a max page limit (15, 30, 50, or 100), and a depth limit (1, 2, or 3 levels). The crawler starts at the seed, fetches its HTML, extracts every same-host internal link, and follows them breadth-first until it hits either the page cap or the depth cap. Links to external domains are ignored. Fragment-only links (#section) are canonicalized away. Trailing slashes are normalized so /page and /page/ don't count as separate nodes.
The Netlify function that does the fetching respects a 10-second per-page timeout. A 30-page crawl typically finishes in about 45 seconds on a normal site.
Reading the Graph
The snapshot card gives you the five numbers: pages crawled, hub count, orphan count, dead-end count, and average in-link count. The average in-link is a crude measure of overall internal-link density. Sites with a thoughtful structure tend to land at 2–4. Below 1 means the site is sparsely interconnected. Above 5 means nav bars and footers are dominating, which can mask real topical linking.
The graph card shows a circular SVG plot of the top 40 nodes by in-degree. The seed is centered and rendered in teal. Hubs are green. Orphans are red. Everything else is gold. Circle size scales with in-degree. Lines between circles represent outbound internal links.
The layout is deliberately simple — circular with depth-based radius — because force-directed layouts in SVG are computationally expensive and produce different results every run. The circular layout is deterministic and fast.
The full ranking table lists every crawled page with its in-degree, out-degree, depth from seed, and flag tags (seed, hub, orphan, dead-end, noindex). Sort mentally by in-degree descending and read the top 20 — those are the pages doing the most authority work on your site.
The orphans card lists every orphan with its title and a link. Fix these by adding at least one inbound internal link from a relevant hub or topic cluster.
The dead ends card lists pages with no internal outbound links. Fix these by adding 2–3 related-content links at the bottom.
When to Run It
Quarterly on your own site. Orphan and dead-end counts drift over time as content gets published and older content gets deprioritized in nav menus. Run a clean crawl every quarter and fix the orphans that matter.
Before a content calendar planning session. If you know which pages are orphans, you can assign "link to X from Y" as a task line item on your content brief rather than remembering after publication.
When you're auditing a competitor's site topology. Run the crawl on a competitor's domain to see how they structure internal linking. Well-ranking sites usually have a recognizable hub-and-spoke pattern; you can copy it without copying their content.
Before migrating to a new CMS or URL structure. A crawl before and a crawl after gives you a before/after diff of internal topology — if the count of orphans tripled after migration, you know redirects or menu links dropped.
Honest Limits
The crawler is breadth-first with hard page and depth caps. On a large site, a 100-page crawl is a sample, not a census. For a full site audit on sites over ~200 pages, a paid crawler with more memory and a longer runtime is still the right tool.
The crawler does not execute JavaScript. Pages where the internal-link menu is rendered by JS after first paint will look link-less. In practice most sites render their primary navigation server-side.
No PageRank or authority-score calculation. In-degree is a coarser signal than weighted authority flow, but it's the one that reliably correlates with "Google sees this page as important relative to the rest of the site."
No external link analysis. The tool is internal-only by design.
How This Fits the Methodology
Internal linking is the cheapest SEO lever with a professional toolchain attached. The $100 Network Chapter 20 (Cross-Domain Linking) models authority flow between your properties; this tool applies the same authority model within a single site. The $20 Agency pillar-cluster chapter (9) predicts that every pillar should be a hub in your internal graph — this tool lets you verify whether the topology matches the theory. And the orphan-and-dead-end hunt is exactly the kind of low-cost, high-leverage fix $97 Launch Chapter 23 puts first for a new site that needs to rank quickly.