What Is a Nofollow Link? From Directive to Hint

Jul 20, 2026 · SEO Glossary

Some SEO folklore does real damage, and "nofollow links are worthless" is high on the list. I've watched people decline a mention on a major industry publication because the link would be nofollowed — turning down qualified traffic, brand exposure, and the citations that follow coverage, all to honor a rule Google itself softened years ago. Time to update the folklore.

What is a nofollow link? It's a link carrying the attribute rel="nofollow" — a signal to search engines saying "I'm not vouching for this destination; don't treat this link as my endorsement." In HTML it looks like: <a href="https://example.com" rel="nofollow">example</a>. Google introduced it in 2005 as a weapon against comment spam, which was then devouring every blog with a form field. The idea: if spammed links passed no credit, spamming would stop paying. (Spammers, being spammers, didn't fully get the memo, but the attribute stuck.)

What Is a Nofollow Link Today? The 2019 Plot Twist

For fourteen years, nofollow was a directive: Google flatly didn't count those links. In September 2019, Google reclassified nofollow as a hint — something its systems may consider or ignore when deciding how to treat a link — for ranking purposes, with crawling and indexing following in March 2020. Simultaneously, Google introduced two sibling attributes to make the intent more precise:

The practical upshot of "hint": Google reserves the right to count a nofollowed link when its systems judge the link genuine — a nofollowed citation from a major news site plausibly carries something now. You can't verify or control it, so the sane planning stance hasn't changed: assume nofollow passes no link equity, and treat anything more as an undisclosed bonus.

So Do Nofollow Links Help SEO or Not?

The honest answer has three layers:

• Directly: probably little, possibly something since 2020, unverifiable either way.
• Indirectly: plenty. A nofollow link from a page your customers read sends real visitors, and real visitors produce the downstream stuff that does move rankings — brand searches, mentions, and followed links from people who discovered you through the nofollowed one. Coverage compounds; the attribute on the first link doesn't stop that.
• Profile-wide: a natural backlink profile contains nofollow links, because the real web nofollows liberally — big publishers, Wikipedia, most forums. A profile that's 100% followed links looks curated, and not in a flattering way.

My working rule: evaluate a link opportunity by whether the audience is real and relevant. If yes, the rel attribute is a footnote. If the only value was the equity and it's nofollowed — it was probably a bad link anyway. The corollary holds too: chasing followed links from junk pages because "at least they pass equity" gets the analysis exactly backwards.

When You Should Nofollow (or Better, Use the Specific Attribute)

1. Anything paid: sponsorships, affiliate links, paid placements — use rel="sponsored". This one's about self-protection: selling followed links violates Google's spam policies and is manual-action territory for the seller.
2. User-generated links: comments, forum threads, reviews — rel="ugc" (or nofollow; both work, and combining them is fine).
3. Links you can't vouch for: the general catch-all case nofollow was born for.

And the anti-list: do not nofollow your own internal links. The PageRank-sculpting era is long dead, and under the current model a nofollowed internal link just discards equity rather than redirecting it. There's no scenario where hoarding-by-nofollow helps; if you don't want a page indexed, that's a noindex job, not a nofollow job. The two get confused constantly and do completely different things.

Auditing the Nofollows You Didn't Know You Had

One more variant worth knowing about: page-level nofollow. A <meta name="robots" content="nofollow"> tag in the head applies the hint to every link on that page at once, which is almost never what anyone wants and usually arrives via a mysterious checkbox in an SEO plugin. I once found it live across a site's entire blog for two years — every internal link in every post, discounted, while the team wondered why deeper pages wouldn't rank. If you take one action after reading this, make it a quick template check for that tag.

This is the part that actually bites sites: accidental nofollow at scale. Plugins that nofollow all external links (and sometimes catch internal ones), themes that stamp nofollow on menu links, a template someone edited in 2018 — I've found sites where every link in the main navigation was nofollowed sitewide, quietly starving deeper pages of internal equity for years. Nobody notices because the links work; browsers don't care about rel attributes.

The check takes minutes with the right lens: crawl your site and filter for internal links carrying nofollow — any hits are almost certainly unintentional and worth fixing the same day. While you're in there, confirm your paid and affiliate links carry sponsored, because that's the direction with actual penalty risk. The nofollow attribute is nine characters long; it earns a surprising share of audit findings for its size.