What does nofollow do to a link on your own site?
On your own site, nofollow only costs you. People used to nofollow their unimportant links so the important ones got more. That trick stopped working around 2008. Matt Cutts, then head of Google's webspam team, showed why in June 2009 with a page worth ten points and ten links, five of them nofollowed. Each followed link used to get two points. Since the change it gets one. His advice still holds: "let PageRank flow freely within your site."
It doesn't keep a page out of Google either. Since 1 March 2020 nofollow is only a hint for crawling and indexing, and using it to block indexing "was never recommended". Google can still find the page through your sitemap or other links.
What to use instead
For your own URLs, Google's page on link attributes points to robots.txt and noindex:
- Out of Google (login, account and thank-you pages):
noindex, on a page Google may crawl so it sees the tag. - Not crawled (cart, internal search, endless filter combinations): a
Disallowin robots.txt. - Sorted or duplicate versions of one page get a canonical to the main one.
- A page that isn't worth the link? Link to it less, or drop the link.
When a nofollow on an internal link is fine
Google documents one case: links to filtered results in faceted navigation, as long as "every anchor pointing to a specific URL" has it (Google, December 2025).
The second comes from software. WordPress puts rel="nofollow" on the comment Reply, "Log in to Reply" and "Cancel reply" links, and a theme may do it on its login link. Leave them: nothing needs to rank on a Reply link.
Your audit flags "internal links with nofollow": what now?
Find the one link behind the number, then decide by the page it points to. A Semrush user on the Shopify forum saw 2,832 of them. Another store owner in the same thread found where theirs came from: "100% of them are the link, on all pages, for my customer account login." Open the source of one flagged page and search for nofollow. A link that turns up on every flagged page sits in the template, and one change there fixes them all.
In MarketingTracer it's the issue "Internal links with nofollow" under SEO audit > Issues. It lists the pages that have it, with the most clicks from Google first when Search Console is connected. After the fix, "Scan again" on a page says "Now fine" or "Not fixed yet".
Demo account, sample data.
If the link points to a page that should rank, like a category or a product, take the nofollow out. A login page or the cart can keep it.
How many internal nofollow links are normal?
Zero, apart from what software like WordPress adds. On 29 October 2026 I crawled every page in the sitemaps of the five sites I run: 1,634 pages with 149,659 links to their own pages, and not one with a nofollow. Two of them have a login button, and one runs in ten languages. Their only nofollow links point at other sites.
If you don't want a link followed, why is it there? Don't bother shaping link value around the page; it doesn't work. Maybe your nofollow links are filter links. Those you handle differently: that's a form, not a link.