Why did Google say the disavow tool "does have effects"?

What did Google actually say about the disavow tool?

Google currently says the disavow tool still works, and John Mueller’s latest comment was aimed directly at people calling it useless.

The comment came from a Reddit thread started by a site owner who said they had disavowed more than 800 supposedly toxic domains, including some guest-post links they had previously bought through Fiverr. Around two weeks later, a keyword that had ranked around #3 dropped sharply.

Several commenters immediately told the owner that disavowing links was pointless. Mueller pushed back. He said there was “a lot of hate” for the tool, that it has its “place & time,” and that it “does have effects.”

Mueller was much more cautious about blaming the ranking drop on the disavow file itself. He said disavowing a handful of links would not normally make a site disappear from Search, advised the owner to fix the file, and suggested checking other explanations such as a recent broad spam update or a major technical problem.

That context changes the meaning of the quote. Google was confirming that disavow can affect how links are treated, not announcing that everyone should start cleaning up backlinks again.

Did Google suddenly change its mind about disavowing backlinks?

No. Google’s current disavow guidance is basically the same position it has held for years: the tool can affect rankings, while most websites should never need it.

The confusion comes from how aggressive Google has become when telling ordinary site owners to leave the tool alone. Over the past few years, Mueller has repeatedly called routine disavow work unnecessary and criticized the obsession with cleaning up random spam links. That gradually turned into a broader belief in SEO circles that disavow had become completely ineffective.

Mueller had already added an important qualification earlier this year. Responding to another backlink question, he said that “most sites don't need” a disavow file, while adding that this does not cover every site. He even described the disavow file as “a tool, not a religion.”

Google’s own Search Console documentation currently says the same thing in less colorful language. Google warns that using disavow incorrectly can harm search performance, then says most sites will never need it because Google normally decides which links to trust on its own.

Those two statements fit together perfectly well. A tool can be powerful enough to hurt a site and still be unnecessary for almost everyone.

What Google has said	What it tells us
Penguin became more granular in 2016	Google got better at neutralizing link spam automatically
SpamBrain began detecting link buyers and link-selling sites at scale	Much more artificial link credit can be removed without a disavow file
Mueller spent 2024 and 2025 telling most sites to stop worrying about toxic links	Routine backlink cleanup became increasingly pointless
Mueller said earlier this year that most sites do not need disavow, but some still do	Google had already rejected the idea that the tool was completely obsolete
Mueller now says disavow “does have effects”	The tool still changes something when it is used in the right situation

If Google ignores spammy backlinks automatically, why can disavow still matter?

Google can ignore huge amounts of link spam automatically today, while the disavow file can still remove credit from links Google would otherwise use.

Google has been moving toward automatic neutralization for a long time. When Penguin was integrated into Google's core ranking systems in 2016, Google said it had become more granular and would devalue spam based on spam signals. In 2022, Google expanded SpamBrain so it could identify both sites buying links and sites created to pass artificial links.

Google was unusually explicit about the consequence of that 2022 change. When SpamBrain neutralizes an unnatural backlink, any ranking credit flowing through that link disappears.

The systems have kept getting more aggressive. Earlier this year, Mueller explained that if Google sees a website repeatedly linking out in ways that violate its policies, its systems may simply ignore all links coming from that site because looking for useful value there is no longer worth the effort.

That explains why thousands of horrible-looking backlinks can accumulate without causing any visible problem. Google may already be assigning them effectively zero value.

But Google still currently lists link analysis and PageRank among its core ranking systems. Useful backlinks have not disappeared from ranking. A disavow instruction can therefore matter when it catches one of those useful links rather than the junk Google was already ignoring.

What does Google’s disavow tool actually do to a backlink?

A Google disavow file asks Google to stop using selected backlinks as ranking inputs; the backlinks themselves stay online and can keep appearing in Search Console.

That distinction causes a lot of confusion. Uploading a file does not contact the linking site, delete the link or remove it from Google's Links report.

Google takes the disavow instruction into account when it recrawls pages containing those backlinks. Its current documentation says this process can take several weeks, which is why uploading or editing a file rarely produces a clean overnight before-and-after test.

Mueller clarified the mechanism again in 2025. He explained that Google does not apply a disavow file as an instant filter over its whole index. Google naturally recrawls the sites containing the links and incorporates the instruction as those pages are processed.

Removing entries works through essentially the same cycle. Google has to encounter the links again before useful link credit can start flowing back.

Can disavowing good backlinks really make Google rankings fall?

Yes. If Google is counting a backlink positively today, disavowing it can remove real ranking credit and hurt the page.

We have a useful real-world test from Ahrefs. Patrick Stox deliberately disavowed 3,476 backlinks pointing at three Ahrefs articles, left the file in place for four weeks, and then removed it.

The first article lost roughly 18% of its estimated organic search traffic during the first month. The second lost 13.3% and returned to about 99% of its earlier level after the disavow was removed. The third barely moved during the disavow period, then gained more than 44% after the links were allowed back into Google's calculations.

It was not a laboratory experiment. Competitors moved, Google changed, individual links were recrawled at different times and Ahrefs itself pointed out those limitations. Still, all three pages together produced a much more useful observation than a single ranking anecdote: search visibility generally weakened after link value was removed and improved after it was restored.

Google's own warning points in the same direction. Search Console explicitly says an incorrectly used disavow file can harm a site's performance. There would be little reason for that warning if every disavowed backlink had already been ignored.

Ahrefs test page	What happened after links were disavowed	What happened after the file was removed
Top YouTube Searches	Search traffic fell about 18% in the first month	Part of the loss later recovered
SEO Pricing	Search traffic fell 13.3%	Traffic returned to about 99% of its previous level
Top Bing Searches	Little initial decline	Traffic later jumped by more than 44%

Does every disavowed backlink affect Google rankings?

No. A disavowed backlink only changes rankings when Google was giving that link some useful credit in the first place.

Imagine 5,000 automated spam links pointing at a website. If Google's systems have already classified all 5,000 as worthless, adding those domains to a disavow file has very little left to accomplish.

The situation changes with a legitimate editorial backlink. If Google trusts the page, trusts the link and uses some of the link's information in ranking, disavowing it removes something that was actually helping.

There is also a messy middle ground. Artificial links can still have some value before Google catches them. Google's 2022 link-spam update explicitly warned that rankings could move when SpamBrain found unnatural links and removed the credit they had been passing.

This gives us a much better mental model for disavow today. The number of domains in a file tells us almost nothing by itself. What matters is how much usable ranking credit Google was assigning to those links before the instruction was submitted.

Why do the Fiverr guest posts matter in Google’s latest disavow case?

The Fiverr guest posts make Google’s latest disavow case much more serious because the site owner had deliberately purchased at least some of the backlinks.

Google's spam policies explicitly cover buying links for ranking purposes. Google has also spent years warning about guest posts created primarily to pass optimized links rather than to reach a genuine audience.

The distinction between purchased guest posts and random scraper backlinks is huge. A normal website can accumulate thousands of strange links without ever asking for them. Google knows that and has built its systems around that reality.

A backlink campaign someone paid for tells Google something different about intent. Some of those links may already have been neutralized. Others may still have been passing ranking credit. A large enough pattern can also create manual-action risk.

That produces an uncomfortable possibility in the Reddit case. If some Fiverr links were still helping the site rank, correctly disavowing them could make rankings worse. The ranking loss would represent artificial credit disappearing rather than the disavow tool malfunctioning.

Google described exactly that behavior when it launched its SpamBrain link-spam improvements. Once artificial link credit is neutralized, the previous ranking advantage disappears with it.

Did the disavow file actually cause the Reddit site’s ranking crash?

We cannot prove that the disavow file caused the Reddit ranking drop, and the latest search data gives the recent spam update a serious competing explanation.

The timing certainly makes disavow worth investigating. The site owner reported a large ranking loss around two weeks after uploading the file, and Google's documentation says disavow changes can take weeks as linking pages are recrawled.

But Google had also just completed a global spam update affecting every language. Google's Search Status Dashboard shows that rollout lasted roughly two days and sixteen hours.

It was not a quiet update either. Search Engine Land later published data from SE Ranking comparing search volatility with a normal period. During the spam update, 16.71% of URLs that had been ranking in Google's top 10 fell beyond the top 100 for the same keyword. The comparable baseline was 9.2%. In that dataset, a top-10 URL was around 1.8 times more likely to disappear beyond position 100 during the update.

That is a large enough disturbance to make a simple “I uploaded disavow, then my ranking fell” timeline unreliable.

Mueller also told the site owner to check for very large technical failures. We do not have the site's complete query history, exact disavow file, crawl data or Search Console information. So the conclusion has to stay narrow: the disavow file could have contributed, but the Reddit thread does not prove that it did.

Are “toxic backlinks” actually a Google concept?

No. Google has rules against link spam, but Google does not use the third-party “toxicity scores” shown in SEO backlink-audit tools.

This distinction is one of the biggest reasons people get disavow wrong.

Google cares about behaviors such as buying links, large-scale link exchanges, automated link creation and guest posts designed primarily to manipulate rankings. A backlink tool has to infer risk from data it can observe itself, such as domain quality, anchor text, traffic patterns and other characteristics.

Those systems cannot see Google's internal classification of a link.

Mueller has been especially blunt about this over the past few years. He has criticized the idea that site owners should routinely export backlinks labeled “toxic” by an SEO tool and feed them into Google's disavow file.

That workflow creates a nasty false-positive problem. A third-party tool only has to mistakenly classify one genuinely useful domain for a site owner to remove ranking value Google was perfectly happy to count.

We would care far more about how a backlink was acquired than whether an SEO platform paints it red.

Can competitors still hurt you with negative-SEO backlinks today?

For normal spam blasts, negative SEO through backlinks is much harder to make work today than many site owners still fear.

Google has spent roughly a decade reducing the ability of bad links to damage innocent sites. Penguin became more granular in 2016. SpamBrain later learned to identify sites buying links and sites created to sell or pass them. Earlier this year, Mueller said Google can even ignore outgoing links from an entire site when the linking behavior looks consistently untrustworthy.

Google's current disavow documentation also says it works hard to stop actions on third-party sites from hurting another website.

The economics make sense. If a competitor could reliably destroy your rankings by spending $50 on 10,000 obvious spam backlinks, Google Search would be extremely easy to sabotage. Cheap, automated links are also among the easiest patterns for Google's spam systems to encounter repeatedly across the web.

Backlink-based negative SEO is not something we would call literally impossible. Google's own documentation leaves room for incoming links to affect its view of a site in some circumstances.

For the ordinary flood of scraper domains, forum spam and auto-generated backlinks that established sites receive these days, though, constantly disavowing them is a solution to a problem Google usually handles itself.

When should you actually use Google’s disavow tool now?

Google currently wants the disavow tool used mainly when a site has a substantial set of artificial links it created or paid for and a manual action, or a realistic risk of getting one.

Google's live Search Console guidance gives two conditions. There should be a considerable number of spammy, artificial or low-quality links, and those links should have caused a manual action or be likely to cause one.

The manual-action documentation makes the use case even clearer. When Google issues an “unnatural links to your site” action, the site owner is told to identify violating backlinks, try to remove or qualify them, disavow the links that cannot be removed, and then request reconsideration.

That explains why Google still keeps the disavow tool around even as Mueller repeatedly tells most people to ignore it. Manual link actions still exist, Google still documents disavow as part of their cleanup process, and some businesses genuinely have years of paid or manipulated backlinks to unwind.

Earlier this year, Mueller captured the current position well when he said most sites do not need a disavow file, but that does not cover every site.

Backlink situation	Should we normally disavow?	Why
Random scraper and junk links we never created	No	Google usually handles these automatically
An SEO tool suddenly reports hundreds of “toxic” domains	No	A third-party toxicity score does not tell us how Google treats those links
A site has a history of buying ranking links	Possibly	The backlink profile may contain links that genuinely violate Google's spam policies
Search Console shows an unnatural-links manual action	Yes, as part of the cleanup	Google explicitly includes disavow in its documented recovery process
Legitimate backlinks were accidentally added to a disavow file	Correct the file	Useful ranking credit may have been removed

Is a manual link penalty different from an algorithmic spam hit?

Yes. A Google manual action is a human-reviewed violation with a Search Console notice, while an algorithmic spam hit can change rankings without any manual-action message.

The difference is important because disavow has a very clear role in one case and a much less obvious role in the other.

With an unnatural-links manual action, Google tells the site owner what category of violation it found. Search Console provides a reconsideration process, and Google's instructions explicitly tell owners to remove or disavow artificial inbound links.

Algorithmic spam systems work continuously. SpamBrain can recognize patterns, neutralize links and change the ranking credit being passed without sending the site owner an unnatural-links manual action.

Google has also explained that a decline during a spam update can happen because its systems start discounting spam links that previously benefited a site. The site loses rankings because some of the old ranking input has disappeared.

Knowing which situation we are dealing with changes the diagnosis completely. A Search Console manual action gives us a specific cleanup problem. A ranking decline that happens near a spam update can have many causes, and a backlink export alone cannot tell us which one applies.

Can disavow fix a Google spam-update drop?

Usually no. If Google's spam systems have stopped counting artificial link credit, disavowing those links cannot restore the ranking benefit Google removed.

Google said this directly when it expanded SpamBrain's link detection. Rankings could change as spammy backlinks were neutralized and the credit passed through those links disappeared.

Suppose a site had 100 purchased guest-post links helping a commercial page. Google then recognizes 70 as link spam and stops counting them. Uploading those same 70 links to a disavow file afterward does not recreate their former ranking power. Google had already decided the credit should disappear.

A broader spam update is even harder to diagnose from backlinks alone. Google's spam policies cover far more than link manipulation, and a site can move during one of these updates for reasons unrelated to inbound links.

This is why we would be very careful with a site that loses traffic around a spam update and immediately receives an automated backlink report telling it to disavow 2,000 domains. The timing of the ranking loss does not establish the cause, and the cleanup can remove good links on top of whatever Google already changed.

Can removing a bad disavow file bring Google rankings back?

Yes, sometimes. If a disavow file blocked backlinks Google still considers useful, removing those entries can let their ranking credit return after Google recrawls the linking pages.

The Ahrefs experiment gives us one example. As seen above, the SEO Pricing page lost 13.3% of its estimated traffic while its backlinks were disavowed and returned to roughly 99% of its previous level after the file was removed.

Google has also said explicitly that removing links from a disavow file tells its systems to consider those backlinks again when they are recrawled.

That does not guarantee recovery. A backlink may have disappeared, lost value or been independently classified as spam since the file was created. Rankings and competitors may also have changed in the meantime.

Old disavow files deserve particular attention. In one Google Search Central discussion, a site owner said they had accumulated more than 11,000 disavowed links over 15 years despite never buying links or participating in link schemes. Mueller told them that if they were sure the file had never been needed for a manual action, he would save a backup, delete it and move on.

That advice should not be copied blindly by a site with a history of paid guest posts or link networks. We would first ask why the file exists. A decade of random internet junk and a decade of deliberate link building may produce similarly large text files while representing completely different risks.

Why did Google say the disavow tool “does have effects”?

Google said the disavow tool “does have effects” because the SEO conversation had drifted too far from “most sites do not need it” toward “the tool cannot change anything,” and the evidence does not support that stronger claim.

The disavow tool can clearly change link signals. Google's current documentation warns that bad use can hurt search performance. Google's ranking documentation still says link analysis and PageRank are part of its core systems. Controlled testing from Ahrefs has shown search visibility falling after useful links were disavowed and improving after they were restored. Google's manual-action instructions still tell some penalized sites to use the tool.

At the same time, Google's automatic link-spam detection is far better than it was when disavow launched. SpamBrain can neutralize artificial links at scale, Google can disregard links from untrustworthy sites, and ordinary websites currently have little reason to spend their time cleaning random backlink spam.

The practical answer is sharp: disavow still works, but using it because Ahrefs, Semrush or another SEO tool found scary-looking domains is usually a bad idea.

The people who should care most about disavow are the ones who know how the suspicious backlinks got there: they bought them, built them, inherited an aggressive SEO campaign or received an unnatural-links manual action.

For everyone else, Google's latest statement changes much less than the headline suggests. The tool still has an effect. The reason Google keeps telling people to leave it alone is that those effects can go in the wrong direction.