Internal Linking

Internal linking is the set of links between pages on your own domain. It is the only part of the link graph you fully control, and it does three jobs at once: it makes pages discoverable to a crawler that has no other route to them, it passes authority from pages that have it to pages that need it, and its anchor text is a plain statement about what the destination page is about. The anchor text is the part most often wasted. Every link that reads read more or click here is a description discarded, and every link that reads exactly the phrase the target page is trying to own is a small, repeated vote for that association. Vary the wording enough to stay readable, and let it describe the destination rather than the act of clicking. Structurally, the pages you care about should be reachable in a small number of clicks from somewhere important, and a cluster of related pages should link to each other rather than all pointing back at a hub. Orphans are the classic failure: a page that exists in the sitemap and in nothing else has been published in a technical sense and connected to nothing, and it will behave accordingly. Automation is where this goes wrong at scale. A script that inserts a link every time a term appears produces thousands of identical anchors, links from contexts where the connection makes no sense, and eventually two pages linking to each other with the same phrase while competing for the same query. Automated linking needs a cap per page, a sense of relevance, and a human reading a sample of the output. The check that finds most problems needs no tooling: crawl your own site and list the pages with no inbound internal links, then list the pages with the most. If the first list contains anything you meant to rank and the second is entirely navigation, the structure is not saying what you think it says.