Stop Adding rel=nofollow to Internal Links

rel="nofollow" is a link attribute, <a href="/page" rel="nofollow">, that tells search engines not to pass ranking credit through that link and, historically, not to bother crawling it. Google introduced it in 2005 specifically to fight blog-comment spam, so people couldn't boost a page's authority just by dropping links into open comment sections.

This audit fires when a large share of a site's internal links, links from one of your own pages to another, carry that attribute. That's different from nofollow on outbound links to other people's sites, which is a separate and often legitimate practice (paid and affiliate links are supposed to use rel="sponsored" specifically, a related but distinct value introduced in 2019).

Excessive internal nofollow usually traces back to one of a few habits: a leftover "PageRank sculpting" strategy from the late 2000s, a page-builder or theme setting that defaults new menu items or buttons to nofollow, a plugin that blanket-applies nofollow to some link pattern without checking whether the destination is actually a real content page, or a mistaken belief that nofollow will keep a page out of Google's index. It won't reliably do that, and it isn't the right tool for it.

This is a warning, not an error, because nofollow on a genuine utility link, login, cart, account settings, is completely fine and often correct. The issue is specifically when it lands on links to pages you actually want found and ranked: blog posts, product pages, category pages, or anything in your main navigation.

The idea that nofollowing some internal links redirects their "share" of PageRank to your other, followed links is a myth, and has been since 2009. Google's own explanation at the time was that a nofollowed link still counts toward the total number of outbound links considered when dividing up a page's PageRank; the portion that would have flowed through the nofollowed link isn't rerouted anywhere, it's simply lost. So the classic sculpting setup, nofollowing low-value pages to concentrate more equity on money pages, never worked the way people thought, even at the time it was popular.

In 2019, Google made things a step more nuanced by changing nofollow from a hard directive into a hint. Since then, Google has said it may choose to crawl a nofollowed link anyway and can take it into account as one signal among many, rather than guaranteeing it's ignored entirely. Practically, this means internal nofollow no longer reliably accomplishes its other historical purpose of keeping a link from being crawled either. You end up with the worst of both outcomes: link equity that isn't cleanly redirected anywhere, and no dependable guarantee the link goes uncrawled.

The real, current cost of nofollowing internal links to real content is crawl discovery. On a large site with a constrained crawl budget, or for a page with no other followed internal links pointing to it, a nofollowed link can be the only path Google has to find that page, and treating it as a hint to skip means slower discovery or, in the worst case, a page that never gets crawled through normal internal linking at all. That compounds with orphan-page risk: a page reachable only through nofollowed links is functionally close to an orphan from a crawler's point of view.

For AI answer engines, internal links, nofollowed or not, matter less directly since most don't compute anything resembling PageRank. But the broader lesson still applies: if you don't want a page crawled, indexed, or cited, nofollow is not a dependable access-control mechanism for any crawler, human-facing search engine or AI bot alike. Use a meta robots noindex tag, or actually restrict access, if that's the real goal.