Most advice on nofollow links is a lightly edited version of Google’s 2019 announcement.
That announcement changed something important, but it did not settle every argument. Google moved nofollow away from being an absolute directive and began treating it as a hint. The practical consequence is easy to miss: nobody outside Google can verify whether a particular nofollow link influenced ranking, crawling or discovery in a specific case.
An article that says nofollow definitely passes ranking value, or definitely never contributes anything, claims more certainty than the evidence allows.
So the useful question around nofollow link value in 2026 is not simply, “Does this link pass equity?”
It is:
What is this placement worth if its ranking value is zero?
That is a much better question for an SEO manager deciding where to spend budget.
This guide uses Google’s documentation, the history of the attribute and practical tests you can run against your own backlink profile to answer it. If you need the basic mechanics first, our guide to dofollow vs nofollow backlinks covers the distinction.
One disclosure before we begin: Infinity Rank sells link-building services, including campaigns where publishers may place followed editorial links. We do not treat compensated links as exempt from Google’s rules. Any paid or sponsored relationship should be qualified in line with Google’s link-spam policies.
The short answer
Do nofollow links help SEO? Yes, but you should not budget for them as measurable ranking equity.
Google says links marked nofollow, sponsored or ugc are generally treated differently from ordinary editorial links, and the attributes function as hints rather than absolute commands.
That still leaves several useful jobs for a nofollow link.
It can send qualified visitors.
It can expose your brand to journalists, creators and potential customers.
It can put your name on platforms that later appear in search results or AI-generated answers.
And it can correctly qualify paid or user-generated links so they do not create unnecessary spam-policy risk.
What nofollow links do not do is give you a dependable unit of ranking value you can plug into a forecast.
In a September 2026 survey reported by PPC Land, 131 SEO practitioners ranked backlinks second among 103 factors they believed mattered for Google rankings. That is evidence of practitioner opinion, not a measurement of Google’s algorithm, but it is consistent with the long-standing importance of backlinks in competitive search markets.
Our working rule is simpler:
Build or pay for a nofollow placement only when it would still make sense with its ranking value set to zero.
That keeps the decision inside your link building ROI model instead of inside a debate nobody outside Google can conclusively settle.
What Google says about nofollow in 2026
A nofollow link is an ordinary hyperlink containing rel="nofollow".
Visitors see the same clickable link. Search engines see an additional instruction describing how the publisher wants the relationship treated.
There is no official dofollow attribute. “Dofollow” is industry shorthand for a normal hyperlink that does not carry one of Google’s qualifying rel values.
Google’s documentation on qualifying outbound links recognizes three relevant values:
rel="sponsored"for advertisements, sponsorships and other compensated placements. This is the preferred value for sponsored links.rel="ugc"for links inside user-generated content such as comments and forum posts.rel="nofollow"for other cases where you do not want to associate your site with the linked page in the normal way.
Values can also be combined, such as rel="ugc nofollow“.
Google announced on September 10, 2019 that nofollow, sponsored and ugc would be treated as hints for ranking purposes.
It separately said that, from March 1, 2020, nofollow would also become a hint for crawling and indexing rather than an absolute instruction.
That distinction matters.
Google’s current documentation says these qualifying values are generally not used in the same way as ordinary links, but a nofollowed destination can still be discovered through other links, sitemaps and crawling paths.
Two similar-looking concepts cause unnecessary confusion.
A robots meta tag containing nofollow applies to links across the page. An anchor-level rel="nofollow" applies only to the individual hyperlink.
And browser-level values such as those covered in our guide to noopener and noreferrer deal with browser security and referrer information. They do not mean the same thing as nofollow for search.
Nofollow is also only one characteristic among many types of backlinks.
A nofollow link from a publication your customers actually read can be more commercially valuable than a followed link buried on an irrelevant directory page.
The adoption data reinforces another point.
The HTTP Archive’s 2025 Web Almanac reported nofollow on 32.3% of pages, while sponsored and ugc each appeared on only 0.5%.
That does not prove why any individual link received its attribute.
It does show that broad nofollow remains dramatically more common than the newer, more specific values. When you audit a backlink, do not assume the attribute reflects a deliberate editorial judgment. Check the type of site and how it applies link attributes across its platform.
How the confusion started
Google introduced nofollow on January 18, 2005 as a response to comment spam.
Matt Cutts and Blogger’s Jason Shellen had been involved in proposing the approach, and blogging platforms quickly adopted it.
English Wikipedia began nofollowing external links in January 2007. That is why Wikipedia backlinks are normally pursued for visibility, credibility and discovery rather than dependable PageRank transfer.
Then came PageRank sculpting.
Some SEOs started adding nofollow to internal links because they believed PageRank that would have flowed through those links could instead be redirected toward more important pages.
In June 2009, Matt Cutts explained that this did not work the way people hoped.
If a page had ten outgoing links and five were nofollowed, PageRank was still divided as though all ten existed. The portions associated with the nofollowed links were not redistributed to the other five.
That largely killed nofollow-based PageRank sculpting.
The 2019 update introduced sponsored and ugc, giving publishers more specific ways to describe outbound links.
None of those changes eliminated link abuse.
Operators using link farms and other manipulative systems simply moved toward tactics where attributes were irrelevant or where publishers would still provide ordinary links.
In January 2026, John Mueller again addressed concerns about spam links left in comment sections, explaining that this kind of spam is not something site owners normally need to panic about for rankings.
The pattern across 2005, 2009 and 2019 is useful.
Each technical clarification was followed by a new industry theory.
First, nofollow was treated like a penalty.
Then it became a supposed PageRank sculpting tool.
Now some marketers talk about nofollow as though Google quietly gives it full-strength ranking value.
There is no good evidence for any of those extremes.
What can be proven and what cannot
Google’s documented position is that qualified links such as nofollow, sponsored and ugc are generally not treated like ordinary editorial links.
Because Google describes these attributes as hints, however, the company retains discretion in how it processes them.
An outsider cannot see that decision for a particular URL.
That means three statements hold up better than most claims in the nofollow debate.
First, nofollow is not a quality penalty.
Martin Splitt explained in 2024 that widespread use of nofollow does not itself tell Google that a site is low quality.
Second, nofollow cannot sensibly be budgeted as a guaranteed ranking input.
If your financial case for a placement requires assumed PageRank transfer, your model depends on something you cannot measure.
Third, nobody outside Google can prove that Google never uses information associated with a nofollow link in any circumstance.
That is the consequence of the hint model.
Be especially careful with studies claiming to prove that nofollow links cause ranking improvements.
Strong brands naturally accumulate both followed and nofollow links.
If high-ranking brands have more nofollow links, that does not mean the nofollow links caused the rankings. Brand popularity may have caused both.
That reverse-causality problem is rarely discussed.
There are also plenty of situations where pages can rank without backlinks, particularly for low-competition queries.
Competitive markets are different. In those cases, genuine editorial links still matter, but that does not turn every followed link into a good link.
Third-party metrics create another source of confusion. Measures discussed in our comparison of domain rating vs domain authority use proprietary link indexes and formulas. They cannot tell you whether Google credited a particular nofollow link.
And do not start disavowing toxic backlinks simply because you see ugly nofollow links from forums or comment sections.
Disavow decisions should focus on manipulative patterns you can actually trace, particularly links associated with schemes, paid networks or deliberate attempts to influence rankings.
The four jobs a nofollow link can do
Instead of asking whether a nofollow link passes equity, ask what job the placement performs.
We use four:
| Job | What the link does | What to measure |
| Discovery | Sends relevant people to your site | Referral traffic, engaged sessions |
| Corroboration | Connects your brand with a topic across independent sources | Brand mentions, citation frequency |
| Conversion | Introduces people who may later mention or link to you | Leads, earned links, outreach responses |
| Insurance | Correctly qualifies paid or user-generated links | Policy compliance, reduced link-spam risk |
A single link can perform more than one job, but you should still be able to name the main one before approving the placement.
Discovery
A nofollow link in a roundup, community thread, profile or publication can send genuine visitors.
Those visitors can be valuable regardless of what happens inside Google’s ranking systems.
Much of what people describe when discussing how social media backlinks improve SEO works this way.
Attention comes first.
Search value may follow later through branded searches, citations, mentions or ordinary editorial links.
Corroboration
Independent sources that repeatedly associate your brand with a topic create public evidence about what your company does.
That does not mean every mention becomes a ranking factor.
It means those mentions exist across the web and can be encountered by people, search systems and AI tools.
Tracking real brand mentions helps you identify where that evidence is appearing.
Google’s own guidance on generative AI features continues to emphasize useful, original and trustworthy content rather than manufactured mentions.
So the distinction matters: corroboration should be earned, not mass-produced.
Conversion
A nofollow link can introduce your business to someone who later links normally.
Journalists, newsletter writers, bloggers, researchers and podcast hosts all browse sources that frequently use nofollow.
That is one reason tactics such as HARO link building and earning podcast backlinks can have effects beyond the first link.
The first exposure does not need to pass PageRank to create a second opportunity.
And when you find an existing mention without a link, link reclamation gives you a straightforward way to ask whether the publisher is willing to add one.
Insurance
Some nofollow-style links have a completely different purpose.
They keep paid, sponsored or user-generated links properly qualified.
There is no ranking benefit to celebrate here, because ranking benefit is not the objective.
The objective is avoiding a link scheme.
Google’s spam policies apply across Search, including surfaces connected with AI Overviews and AI Mode.
So qualifying paid relationships is not wasted link equity. It is basic risk control.
A placement that cannot justify itself through discovery, corroboration, conversion or insurance is probably a link bought on hope.
Do nofollow links matter for AI search?
There is no documented Google rule saying nofollow links receive special treatment inside AI Overviews or AI Mode.
There is also no documented rule saying the attribute becomes irrelevant.
Google’s guide to generative AI features says AI Overviews and AI Mode rely on core Search ranking and quality systems and may use techniques such as retrieval-augmented generation and query fan-out.
Google also says familiar SEO fundamentals still apply.
It does not recommend special AI markup, content chunking or llms.txt for Google Search.
Most importantly for this topic, the guide does not publish a followed-versus-nofollow weighting system for AI-generated results.
Silence is not evidence either way.
Outside Google, some vendor studies attempt to measure citation patterns.
Analyze AI, citing research from Discovered Labs, has reported very high citation shares for platforms such as Wikipedia and Reddit in particular AI products.
Treat those figures as directional, not as universal facts. They come from commercial AI-visibility research, not from the companies operating the search systems themselves.
The useful observation is narrower.
Several platforms that appear frequently in AI citations, including Wikipedia and Reddit, commonly qualify outbound links with nofollow or UGC attributes.
That means work aimed at ranking your website on ChatGPT or applying best practices for answer engine optimization can naturally create useful mentions even when the outgoing link is qualified.
The same applies to entity work.
Our guide to knowledge graph optimization covers how consistent entity information helps connect a brand with the right topics.
Community platforms show the distinction clearly. Quora link building may produce a nofollow link, but the page, answer and brand mention can still be discovered and cited independently of PageRank transfer.
You can test this against your own market.
Write twenty questions your buyers actually ask.
Run them through the AI tools your customers use.
Record every cited domain.
Then inspect the outbound links on those domains.
The percentage of frequently cited sources that use nofollow or UGC in your category is more useful than a generic rule about what an ideal backlink profile should look like.
Nofollow vs dofollow ratio: why the number is the wrong target
Google does not publish an ideal nofollow-to-followed ratio.
Industry recommendations vary wildly.
Some agencies suggest 70/30.
Others suggest 60/40.
Others correctly point out that no universal ratio exists.
The problem is not simply that the percentages disagree.
The denominator itself can be misleading.
Imagine three topic clusters where the nofollow shares are 30%, 90% and 50%.
The sitewide average is about 57%.
That percentage accurately describes the arithmetic and accurately describes none of the three clusters.
Older Semrush research also found substantial variation in nofollow shares among high-ranking pages and sites. That is useful as evidence that successful backlink profiles vary, not as a benchmark you should copy.
Measure three narrower things instead.
Per money page: calculate the nofollow share among referring domains pointing at the specific URL you care about.
Per referring domain: count each domain once. Our explanation of referring domain vs backlink shows why one website linking thousands of times can distort raw backlink totals.
Per topic cluster: compare your profile against three to five genuine search competitors. A backlink gap analysis shows which domains link to competitors but not to you.
You can supplement that with competitor backlinks research and whichever backlink monitoring tools you trust.
Remember that every commercial backlink index is incomplete. If Search Console is not showing backlinks data, do not mistake a third-party crawler’s database for the complete web.
Context matters more than arithmetic.
A lopsided profile made up of genuinely niche relevant backlinks is healthier than a mathematically perfect 60/40 split assembled from irrelevant websites.
Treat your ratio like a smoke alarm, not a thermostat.
A sudden change deserves investigation.
A stable number does not need constant adjustment.
A nofollow link building strategy that earns its keep
Start by naming the job.
For every target in your outreach list, write down whether the placement serves discovery, corroboration, conversion or insurance.
If you cannot name the job, remove the target.
That simple filter makes a better link building checklist than chasing nofollow percentages.
Next, price the audience.
Estimate the qualified visits the placement could send over twelve months.
Multiply that by what a qualified visit is worth to the business.
Then compare the result with the cost.
For example, suppose a roundup could send 300 relevant visitors over a year and a qualified visitor is worth $5 to you.
That gives the audience component an estimated value of $1,500.
Paying $4,000 for that placement only makes sense if there is another measurable job it performs. If the calculation requires assumed PageRank transfer to work, walk away.
Then go where nofollow is native.
Directories, community answers, expert roundups and press distribution often qualify links automatically.
Web directory link building makes the most sense when the directory has real local, industry or category relevance.
Likewise, press release SEO best practices should focus on producing something journalists may actually use, not stuffing a release with anchor text.
Give arriving visitors something worth citing.
Strong linkable assets such as original research, useful calculators, benchmarks and reference guides turn discovery into future linking opportunities.
That is how nofollow work can support the broader acquisition of contextual backlinks and genuine authority link building.
Keep those two budgets conceptually separate.
Nofollow placements should justify themselves through audience, visibility or compliance.
Ranking-focused link acquisition should prioritize genuinely earned editorial endorsements rather than assuming compensation can buy PageRank safely.
Review the results quarterly.
Did the placement send traffic?
Did it create citations?
Did it result in new mentions or links?
Did it serve a compliance purpose?
If the answer to all four is no, remove that source from future link acquisition plans.
Mass profile creation and comment drops fail this test almost immediately.
They rarely provide an audience, credible corroboration or meaningful conversion path.
When nofollow is the right call on your own site
On your own pages, the basic rule is simple.
Use ugc for links submitted through user-generated content.
Use sponsored for advertising, affiliate arrangements and other compensated relationships.
Use nofollow for the occasional link you do not want treated as a normal editorial endorsement.
Do not use nofollow as an internal crawling or indexing strategy.
If you want to prevent crawling, use the appropriate robots controls.
If you want to prevent indexing, use noindex where Google can crawl the page and see it.
Those are different jobs.
Over-nofollowing your own editorial links is another common mistake.
There is normally no reason to nofollow reputable sources simply because they lead away from your site.
The benefits of linking to other sources in blog posts include making your evidence inspectable.
Your own social profiles generally do not need to be nofollowed either.
Template links deserve separate scrutiny.
Sitewide backlinks and footer links can repeat across hundreds or thousands of pages, so one widget credit, partner link or template element can distort backlink counts dramatically.
Guest content requires a policy rather than an automatic answer.
If contributors are allowed to add arbitrary commercial links, qualifying those links is sensible.
If you genuinely edit and approve each citation, make the decision based on the relationship and editorial purpose rather than applying nofollow mechanically.
The same risk difference appears when comparing guest posting vs niche edits.
Finally, remember that a rel attribute is invisible to ordinary readers.
An affiliate or sponsored relationship can therefore be correctly qualified for search engines while still failing consumer-disclosure requirements.
The FTC’s endorsement guidance requires material relationships to be disclosed clearly and conspicuously.
So use the appropriate rel value and visible disclosure.
A surprising number of link building mistakes come from treating the HTML attribute as though it solves every legal, editorial and SEO problem at once.
A 45-minute nofollow audit
You can get a useful picture of a domain in one focused session.
1. Export referring domains: 10 minutes
Pull referring domains from at least two backlink tools plus Google Search Console.
Merge the lists.
Do not start by counting every individual backlink. Start with referring domains so sitewide links do not dominate the picture.
2. Verify the rel values: 10 minutes
Sample ten important links.
Inspect the anchor element in the rendered browser DOM and, where useful, compare it with the raw page source.
Look for nofollow, ugc and sponsored.
Backlink tools can lag behind page changes, and dynamically rendered pages do not always expose the same information in raw source.
3. Tag each domain by job: 10 minutes
Label every sampled domain:
- Discovery
- Corroboration
- Conversion
- Insurance
- Unclear
The unclear group becomes your investigation queue.
4. Flag suspicious followed links: 10 minutes
Look for exact-match commercial anchors, repeated placements across obvious networks and structures associated with link wheels or tiered link building.
Also sample manually for hidden backlinks and placements that commercial crawlers may not classify correctly.
5. Decide: 5 minutes
For each questionable link, choose one action:
Ask the publisher to qualify it.
Remove it if appropriate.
Disavow only when you have evidence of a manipulative pattern that warrants it.
Or leave it alone.
Write the reasoning down.
That record is more useful during the next audit than another exported spreadsheet with no decisions attached.
Where we stand
Infinity Rank sells link-building services.
The service should be judged on the work involved, the quality and relevance of the publishers, the audience a placement can reach and the resulting business value.
We do not recommend valuing a paid nofollow placement as hidden ranking equity.
Likewise, no provider should tell you that compensation makes a followed link automatically safe for ranking purposes.
Google’s spam policies specifically address links created primarily to manipulate rankings, and paid relationships should be qualified appropriately.
When comparing link building pricing, ask the provider what you are actually paying for.
Prospecting?
Content?
Outreach?
Publisher access?
Referral audience?
Brand visibility?
A guaranteed followed link whose value depends on passing PageRank?
Those are not equivalent products.
If you outsource link building, ask the provider to show you the final URL, the actual rel value and the reason each placement belongs in the campaign.
If they cannot explain the job a link performs without talking about DR alone, the strategy is incomplete.
Key takeaways
- Google treats
nofollow,sponsoredandugcas hints rather than simple universal commands. - You cannot reliably forecast the ranking value of an individual nofollow link.
- Judge nofollow placements by four jobs: discovery, corroboration, conversion and insurance.
- There is no official nofollow-to-followed ratio.
- Measure ratios per page, per referring domain and per topic cluster instead of steering toward one sitewide percentage.
- Use natural link building tips to earn genuine editorial citations rather than manufacturing a profile that merely looks balanced.
- Favor manual link building where a person actually reviews the page, audience and link treatment.
- Use
sponsoredandugcwhere appropriate, and do not confuse search-engine markup with reader-facing disclosure.
FAQs
Do nofollow links help SEO?
Yes, but not as a dependable unit of ranking equity. Google treats nofollow, sponsored and ugc as hints, and qualified links are generally not handled like ordinary editorial links. Their practical value comes from traffic, brand discovery, citations and future linking opportunities. Treat them as support for your wider SEO strategy, not guaranteed PageRank.
Does Google ignore nofollow links?
Not completely. Google changed nofollow from a strict directive to a hint for ranking in 2019 and for crawling and indexing in 2020. That gives Google discretion in how it processes individual links. A nofollowed destination can also be discovered through sitemaps, other websites and other crawl paths.
What is a good nofollow to dofollow ratio?
There is no official ideal ratio. Published benchmarks conflict because backlink profiles vary by industry, page type and publisher mix. Measure the ratio per money page, per referring domain and per topic cluster instead, then compare it with competitors and your other link building metrics.
Are nofollow links worth building?
Sometimes. A nofollow link is worth building when the audience, credibility, referral traffic or conversion potential justifies the effort even if no ranking credit transfers. Relevant communities, publications, directories and forum link building can qualify. Mass profile creation and automated comment drops usually do not.
What is the difference between nofollow, sponsored and ugc?
All three are rel values used to qualify links. sponsored identifies paid or advertising relationships, ugc identifies links added through user-generated content, and nofollow covers other links you do not want treated as ordinary endorsements. Google allows values to be combined when more than one description applies.
Do nofollow links help with AI search and ChatGPT citations?
They can still matter for visibility, but no published Google rule gives nofollow links special AI-search weighting. Many platforms frequently cited by AI systems qualify outbound links, so a useful citation or brand mention can exist without a followed backlink. Treat the citation and audience exposure as the value, not assumed PageRank.
Can nofollow links hurt my site?
A normal nofollow backlink is not inherently harmful. The bigger risk is the opposite situation: a compensated or manipulative link left as an ordinary followed link when it should be qualified. Spammy comment links generally do not justify panic. Investigate deliberate link schemes, not random internet noise.
Should I nofollow internal links?
Usually not. Nofollowing internal links to redistribute PageRank has not worked as intended since Google’s 2009 PageRank-sculpting clarification. If you need to block crawling, use robots controls. If you need to prevent indexing, use the appropriate noindex directive instead of trying to solve the problem with internal nofollow.
How can I check whether a link is nofollow?
Inspect the anchor element in your browser and look for rel="nofollow", rel="ugc" or rel="sponsored". Backlink tools and browser extensions can help, but verify important links manually because their databases can lag. On dynamic pages, inspect the rendered DOM rather than relying only on raw page source.
Where do most nofollow backlinks come from?
Common sources include social platforms, community sites, comment systems, directories, press-distribution platforms and reference sites such as Wikipedia. Many free ways to create backlinks for your website also produce qualified links. Judge them by audience quality and relevance, not simply by whether they are followed.
