Nofollow: Did Google really implement its changes on the announced dates?

What really happened with nofollow?

Google has long treated nofollow as a strict directive: no link tracking, no PageRank transfer, end of story. In September and March, Mountain View announced two key dates to transform this behavior. September 1 was meant to mark the transition of nofollow into a potential ranking indicator, while March 1 aimed to integrate it into crawl and indexing decisions.

But here’s the catch: Google never confirmed that these changes were actually activated. We have dates, but zero post-launch communication. No "it’s done", no official feedback, nothing. It’s like announcing an algo update without ever saying if it was really rolled out.

Why does this announcement change the game for SEOs?

Until now, a nofollow link was a guarantee: Google wouldn’t follow it, wouldn’t pass any juice. Handy for sculpting your PageRank, isolating low-value content, or marking sponsored links risk-free. With this announced shift, nofollow becomes an indicator that Google can choose to ignore.

In practical terms? If Google decides that a nofollow link deserves to be followed—because the semantic context is strong, the source page is authoritative, or the overall signal seems relevant—it can do so. The practitioner loses some control they thought they had. And that’s where it gets tricky: there’s no way to know when Google activates this option.

What does this imply for managing internal and external links?

First, PageRank sculpting through nofollow becomes risky. If Google can choose to follow or not follow the link at will, we no longer control the internal juice distribution. Strategies that relied on nofollow to channel flow to priority pages lose predictability.

Next, indexing pages that we wanted to exclude becomes a real risk. A page set to nofollow to keep it out of the index may be crawled and indexed if Google deems the signal strong enough elsewhere. We thought we were holding a firm lever, but now we have a volume knob that Google can turn at will.

• Nofollow is no longer an absolute directive, but an indicator that Google can interpret
• No official confirmation that the changes announced for September and March were actually implemented
• Loss of control over PageRank sculpting and crawl budget management via nofollow
• Risk of indexing pages that we wanted to exclude if other signals contradict the nofollow
• Need to resort to robots.txt or noindex for strict control of indexing

Is this statement consistent with what we observe in the field?

Let's be honest: impossible to decide without concrete data. Google has set two temporal markers, but zero measurable feedback allows us to confirm that the interpretation of nofollow has actually changed on those dates. Field tests show ambiguous behaviors… Some nofollow links seem to be followed at times, others not, without a clear pattern. [To verify]: is this related to the announced evolution or to other factors (semantic context, source authority, content freshness)?

What complicates the analysis is that Google has always had a latitude of interpretation on directives. Nofollow is not the first attribute to go from a "strict command" to a "consultative indicator". The meta robots "noarchive", for example, has never been respected 100% in all contexts. So, noticing discrepancies doesn’t necessarily prove that the announced shift has occurred—just that Google is doing… Google.

What nuances should be added to this announcement?

First point: Google never said, "we activate everything, everywhere, right now". The announcement mentioned "potential" use in ranking algorithms. Potential. This leaves a wide margin for interpretation. Maybe only certain sectors, types of links, or configurations trigger this interpretation. We don’t know.

Second nuance: crawling and indexing are not ranking. Even if Google decides to crawl a page despite a nofollow, it guarantees neither its indexing nor its ranking. It may very well discover the resource, deem it irrelevant, and set it aside. The reverse is also true: a nofollow link may be ignored for crawling but used as a contextual signal for ranking. The mechanisms are not synchronized.

In what cases might this rule not apply?

Google has introduced alternative attributes to nofollow: rel="sponsored" for paid links, rel="ugc" for user-generated content. If Google's goal was to better qualify links rather than block them outright, it’s likely that these more precise attributes are treated differently than generic nofollow. [To verify]: does a rel="ugc" link have a higher chance of being followed than a classic nofollow? No official data confirms this.

Another case: nofollow links on highly authoritative pages. If a page like Wikipedia, the New York Times, or a government institution adds a nofollow, Google may assess that the contextual signal (co-occurrence, semantic proximity, anchor) still has value, even without PageRank transfer. Conversely, a nofollow on a regular site without authority is unlikely to be reevaluated. The weight of the source probably matters—but again, this is speculation.

What should you actually do with your nofollow links?

First action: audit all uses of nofollow on your site. Identify where you use it and why. If it’s to block indexing, switch to noindex or robots.txt. If it’s to avoid passing juice to low-value pages, reconsider the architecture: it might be better to restructure internal linking than to rely on an attribute that has become uncertain.

Second reflex: monitor your coverage reports in Search Console. If pages meant to stay out of the index (thanks to nofollow) suddenly appear in the index, that’s a strong signal that Google is now interpreting these links. A classic case: pagination pages, facet filters, poorly isolated testing environments. React quickly to avoid polluting the index with duplicate or irrelevant content.

What mistakes to avoid with this new reality?

Mistake #1: continue to sculpt your PageRank via nofollow as if nothing has changed. This technique, already not recommended for years, becomes downright counterproductive if Google can choose to follow the link or not. You lose predictability, and Google may interpret your strategy as an attempt to manipulate.

Mistake #2: ignore the sponsored and ugc attributes. Google introduced these qualifiers for a reason: to better understand the context of the link. If you’re still using generic nofollow on sponsored links or user-generated content, you’re missing out on a quality signal that Google may value (or penalize if misused). Be precise in your attributes; it helps the algorithm better interpret your intent.

How to check that your site is compliant with this evolution?

Run a full crawl with Screaming Frog or Oncrawl filtering all internal and external nofollow links. For each occurrence, ask yourself: does this link really need to be nofollow, or is it a remnant of an old strategy? If the answer is unclear, it’s better to remove the attribute and let Google decide naturally.

Next, cross-check with the crawl data from Search Console. Compare crawled pages with those you thought you were excluding via nofollow. Any discrepancy is a red flag. If non-strategic pages appear in the index, it’s because Google decided to follow them despite your directives. Adjust then with noindex or a robots.txt block.

• Audit all internal and external nofollow links, identify their reason for being
• Switch to noindex or robots.txt for strict control over indexing
• Use rel="sponsored" and rel="ugc" instead of generic nofollow when relevant
• Monitor Search Console coverage reports for any unexpected indexing
• Regularly crawl the site to identify pages crawled despite a nofollow
• Review internal linking architecture to avoid depending on nofollow for PageRank distribution