Yes, nofollow links help SEO, but indirectly and less than a good followed link. Since 2019 Google treats nofollow as a hint rather than a hard instruction, so it may pass some ranking value, and separately these links still deliver the things that produce rankings later: qualified traffic, brand mentions, and discovery by people who go on to link to you properly.
The useful version of this question is not whether the attribute helps. It is whether a specific opportunity in front of you is worth an hour of work. That depends on the page it sits on and who reads it, not on what is in the rel attribute, and most of this post is how to make that call quickly.
Last updated 7 September 2026.
What Google actually does with nofollow now
Google announced the change on 10 September 2019. For ranking purposes, nofollow became a hint from that announcement. For crawling and indexing purposes, it became a hint on 1 March 2020. Two new values arrived at the same time: rel="sponsored" for paid or compensated placements, and rel="ugc" for user-generated content like comments and forum posts.
Hint means Google decides. It may follow the link, pass value, or ignore it entirely, and it does not tell you which. Google’s current documentation words this carefully: links with these attributes will “generally not be followed”, while noting the target may still be crawled through other routes. So you can no longer treat nofollow as a switch that turns a link off. It is a signal about the nature of the link, weighed alongside everything else Google knows about both pages. That is the whole technical story, and anyone offering more certainty is guessing.
Two nofollow links I would pursue again
Two situations from actual link building work, described without client names.
Business directory citations for a local service business
Almost every business directory nofollows outbound listings, and people use that as a reason to skip them. On a local consultation client the listing work was one of four things we ran, and it mattered for a reason unrelated to PageRank: consistent name, address and phone data across directories is a verification signal for local search, checked independently of whether those links pass equity.
The measurable outcome there was phone calls and messages, not ranking movement on the directory pages. If you are local, treat citations as a data-consistency job that happens to produce links. The same outreach ran on a reviews and directory platform I worked on for six months, alongside industry publication coverage, written up with the traffic figures in my case studies.
A guest article where the byline turned out nofollow
This happens constantly. You pitch an industry publication, they accept, you write the piece, and the byline publishes with nofollow because the site applies it to every external link by policy. Nobody told you in advance.
I would still take it, for three reasons that survive the attribute. The article puts your name in front of the exact audience you sell to. It gives you something credible to reference in later outreach. And the secondary links matter more than the primary one: readers who find the piece useful cite it from their own sites, and those citations are usually followed. The nofollow byline is the seed, not the harvest.
Pursue it or skip it: a four-point check
Run any nofollow opportunity through these four questions before spending time on it.
- Is the linking page relevant, not just the site? A marketing blog is not relevant to you. An article about the specific problem your page solves is. Judge the page, because that is what Google and a reader both see.
- Does the page get traffic, or sit where people actually go? This is the decisive question for nofollow specifically. Referral clicks are the one benefit the attribute cannot take away, so a nofollow link on a page nobody visits is worth close to nothing.
- Would its readers plausibly buy from you? Audience overlap, not audience size. A hundred readers in your niche beat fifty thousand who will never need what you sell.
- Would you be comfortable if a client saw your brand there? A blunt but reliable filter. It catches link farms, listicles with four hundred outbound links, and directories that exist only to sell placements.
Pursue it if it clears question one plus at least one of two or three. Skip it if it fails question four, no matter how good the rest looks, or if it clears only question one and would take more than about thirty minutes. Relevance alone is not payment.
None of the four questions is about the rel attribute, deliberately. If an opportunity passes this check it is worth doing either way, and if it fails, a dofollow link will not rescue it.
Do AI answer engines care about nofollow?
Worth separating what is documented from what is inference, because this area attracts confident claims nobody can support.
Confirmed: Google’s AI Overviews are grounded in Google Search, so anything nofollow does to your organic rankings carries through indirectly. Beyond that, neither OpenAI, nor Perplexity, nor Google has published guidance saying rel attributes influence which sources get cited. If someone tells you nofollow links do or do not count for AI visibility, they are not quoting documentation, because none exists.
Reasonable inference: these systems work differently from PageRank, which is a property of a link graph computed by following edges between pages. Citation selection in an assistant is a retrieval problem: fetch candidate documents for a query, pick the ones that answer it clearly. A mention of your brand on a page that gets retrieved puts your name into that context whether the link is followed or not, because what is being read is the text.
If that holds, nofollow should matter less for AI visibility than for rankings, and unlinked brand mentions should carry more weight than a link-graph model predicts. I think that is likely, and I have seen AI citations appear on a client before the corresponding rankings did. But I have not run a controlled test isolating the attribute, and nor has anyone else that I have seen. Treat it as a working hypothesis, not a reason to restructure your link building.
Two mistakes worth avoiding
Nofollowing your own internal links. This was a real technique called PageRank sculpting, and Google changed how it worked in 2009. Nofollowing an internal link no longer redistributes that authority elsewhere, it discards it. If you do not want a page to receive internal authority, link to it less. Count and placement are the levers that still work, which I covered in how many internal links per page a site needs.
Auditing the attribute instead of the link. I have watched teams filter a prospect list down to dofollow only and end up chasing worse opportunities than the ones they discarded. The attribute is one weak signal. Page relevance, traffic and audience are three strong ones. Sort by those and check the attribute afterwards, which is how I approach link building for clients.
Frequently asked questions
Do nofollow links help SEO?
Yes, though less than followed links and mostly indirectly. Google has treated nofollow as a hint rather than a directive since 2019, so some value may pass. The reliable benefits are referral traffic, brand exposure to a relevant audience, and the secondary followed links from people who found you through the mention.
Are nofollow links worth building deliberately?
Some are. Local citations and relevant industry mentions are worth pursuing on their own merits. Buying nofollow links or mass-submitting to directories is not, because neither sends traffic. If the opportunity would be worth taking with the attribute removed, take it.
Do social media links count as backlinks?
Not as followed backlinks. Effectively every major platform nofollows outbound links. They still matter for distribution, because a post that reaches the right people produces the visits and follow-on citations that do count. Judge them by traffic and attention, not as entries in a backlink profile.
Should I nofollow internal links?
Almost never. PageRank sculpting stopped working in 2009, and nofollowing an internal link discards the value rather than redirecting it. The rare legitimate case is a link you genuinely do not want associated with your site, such as a user-submitted URL in comments, where rel="ugc" is the better value anyway.
Do affiliate links need the sponsored attribute?
Yes. Any link created through payment, affiliate arrangement or other compensation should carry rel="sponsored". Google still accepts nofollow here as a legacy option but prefers sponsored. This is the one place the attribute genuinely matters, because getting it wrong looks like undisclosed paid linking.
About the author. Kapil Lodhi is an SEO and AI SEO specialist in Indore, India, with over four years of hands-on experience across technical SEO, content strategy, link building, local SEO and e-commerce SEO for clients in India, the United States and Canada. He has built links and citations for local service businesses, e-commerce brands and directory platforms, and stopped sorting prospect lists by rel attribute a long time ago.