Why homepage anchor profiles and deep-page profiles must be read separately

Why homepage anchor profiles and deep-page profiles must be read separately

Auditors often blend a site's entire anchor profile into one distribution. That averaging hides a real structural difference between how homepages and deep pages naturally earn anchors.

The pattern: homepages overwhelmingly attract branded and naked-URL anchors — people cite the brand and link to the front door. Deep pages, by contrast, earn more descriptive and partial-match anchors, because people link to a specific article about something, naming that something. So a healthy deep page can carry a higher keyword-anchor fraction than a healthy homepage without that being a red flag.

On one hand, this means a blended profile can look dangerously keyword-heavy if your deep pages earned legitimate descriptive links — a false alarm from improper aggregation. On the other, it means manipulators sometimes hide exact-match links on the homepage where they're least expected, and blending masks that too.

A 2022 study segmenting by URL depth found homepages skewing branded/naked and deep content pages skewing descriptive — consistent with link-acquisition psychology, though depth correlates with content type, confounding it.

Limitation: "deep page" conflates many content types with different natural anchor norms.

Practical read: audit per-URL-class, not per-domain; the safe distribution differs by page role.

Open question: does Google's over-optimization model implicitly expect different anchor distributions by URL depth or page type?