TL;DR
- Search Engine Roundtable reported the exchange on 2 September 2026 and recalled that Google has said the disavow tool hurts many more sites than it helps, and that Bing has removed its equivalent.
- Mueller wrote that disavowing a handful of links is not going to make a site drop out of search, so a small file cannot explain a large drop.
- Executive editor Barry Schwartz noted the site in question was about buying guest posts and links from Fiverr, so the comment was specific to that case.
- The causes he named were a broader spam update around that time, where recovery can take many months, and technical failures on the scale of blocking the whole site.
On 2 September 2026, Search Engine Roundtable reported that Google's John Mueller told a site owner on Reddit that the disavow tool "does have effects," while adding that "it has its place & time." The thread came from a site owner whose rankings dropped after links were added to the disavow file. Mueller did not tell the owner that disavowing had caused the drop, and he did not tell them the tool was useless. He said the file should be fixed so that it says what the owner actually wants it to say, and that the real cause of the drop was probably somewhere else.
What John Mueller wrote in the Reddit thread
According to Search Engine Roundtable's 2 September 2026 report, Mueller opened by pushing back on the mood of the thread itself: "There's a lot of hate on the disavow tool here - but it has its place & time, it does have effects." That is a narrow statement and it is worth reading it as one. He confirmed the tool is connected to something real inside Google's systems. He did not say it is a routine part of SEO, and he did not say the owner in the thread had been hurt by it.
The second sentence is the one most site owners need. Mueller wrote that "disavowing a handful of links is not going to make your site drop out of search." A short disavow file, in other words, is not a plausible explanation for a site disappearing from results. That sets a boundary on how much blame the file can carry. If a small file cannot sink a site, then a drop that follows a small upload needs a different explanation, and the search for that explanation is where the owner's time should go.
Mueller then gave two pieces of direction. The first was to fix the disavow file so it is what the owner needs it to be, which he said "might be empty, or might not be empty." The second was to look for other potential reasons for the drop. He named a broader spam update that may have happened around the same time, noting that the owner had not given a date in the post and that he did not recall the date of the update himself. He warned that when a drop lines up with a bigger algorithmic change, recovery "can take quite some time - sometimes many months." His other suggestion was to rule out large technical problems, and he drew the line clearly: not "oops, wrong canonical" but more like "oops, blocked the whole site."
The thread was about bought links, which changes how the quote should be read
Search Engine Roundtable added context that matters more than the quote itself. Executive editor Barry Schwartz noted that the site Mueller was answering was about buying guest posts and links from Fiverr, or something like that, and that Mueller's statement was specific to that situation. He also wrote that most of the time, sites should not disavow links, though there are occasions when they should.
So the sentence "it does have effects" was aimed at a site owner who had been paying for links. That is the narrow case where a disavow file has a job to do, because the owner knows exactly which links they bought and can list them. It is not a licence for a site that has never bought a link to start uploading domains it does not like the look of.
Search Engine Roundtable also recalled Google's longer track record on the tool: that Mueller has previously said most sites do not need to disavow and that disavowing links is a waste of time, that Google has said the disavow tool hurts many more sites than it helps and has doubled down on not using it, and that Google may remove the tool at some point, as Bing already has. Those two positions are not in conflict. A tool can do something and still be the wrong choice for almost everyone who reaches for it.
The reported facts in one table
Everything in the table below comes from the Search Engine Roundtable report of 2 September 2026 and the Reddit answer it quotes. Nothing here is inference.
| Point | What Search Engine Roundtable reported |
|---|---|
| Mueller on whether the tool does anything | It "does have effects," but "it has its place & time" |
| Mueller on the scale of harm from a few links | "Disavowing a handful of links is not going to make your site drop out of search" |
| First thing Mueller told the owner to do | Fix the disavow file to be what it needs to be, which "might be empty, or might not be empty" |
| Other causes Mueller raised | A broader spam update around that time, with recovery that "can take quite some time - sometimes many months," and big technical problems on the scale of blocking the whole site |
| Google's earlier position on the tool | Most sites do not need to disavow; the tool hurts many more sites than it helps; Bing has removed its equivalent |
Why a drop that lines up with a spam update is usually not the disavow file
This section is analysis rather than anything Mueller said. Two things happening in the same week is the weakest kind of evidence, and a disavow upload is one of the few SEO actions with a timestamp attached, which is exactly why it attracts blame. The owner remembers the day they uploaded the file. They do not remember the day a spam update finished rolling out, because nobody sends them an email about it.
The asymmetry gets worse when you consider scale. A disavow file acts on the specific links listed in it. A spam update acts on how a whole class of sites is assessed. If a site loses visibility across many unrelated keywords at once, including pages that never had the disavowed links pointing at them, the pattern points at a sitewide reassessment rather than at a handful of neutralised links. If instead the losses are concentrated on the two or three pages that the disavowed links used to point at, the file becomes a more reasonable suspect.
Mueller's warning about timelines matters here as well. He said fixing issues around bigger algorithmic changes can take many months to show an effect. That has a practical consequence for anyone who reacts to a drop by making a large change, waiting three weeks, seeing nothing, and then making another large change. By the time the first change could have shown its effect, the site has been altered four more times and there is no way left to tell which change did what.
When a disavow file earns its place, and when it does not
Framing this as analysis: the disavow file is a confession, not a cleaning product. It is useful when a site owner can name the links they are responsible for, usually because they or a previous supplier paid for them. That was the situation in the Reddit thread. The owner has a list, the list is real, and disavowing it is a way of saying so.
The file is a poor fit for the far more common situation, which is a site owner opening a backlink tool, sorting by whatever metric the tool calls toxicity, and uploading everything below a threshold. Those scores are the tool vendor's opinion. Google has never published a threshold, and Search Engine Roundtable's report repeats Google's position that the tool hurts many more sites than it helps, which is what you would expect if most uploads are lists of links that were doing no harm and some good. Anyone whose profile was built through editorial outreach and digital PR rather than through purchases has very little to confess and very little reason to keep a file at all.
There is a middle case worth naming. Sites that changed hands, or that worked with a supplier whose methods were never documented, sometimes genuinely do not know what is in their profile. The honest answer there is an audit before an upload, not an upload as the audit.
What to check before you touch the disavow file
The following is a practical sequence, not advice attributed to Mueller. He named two categories to investigate, and the checks below sit inside those categories.
- Write down the exact date the visibility fell, from Search Console rather than from memory. Mueller specifically noted the Reddit poster had not given a date, which made his diagnosis harder. Without a date, no theory can be tested.
- Check whether the fall is sitewide or page-level. Compare impressions for the pages the disavowed links pointed at against the rest of the site over the same window.
- Rule out the large technical failures Mueller described, on the scale of blocking the whole site: a robots.txt file that shipped with a disallow rule, a staging noindex tag that reached production, a server returning errors to Googlebot, an expired certificate. These are cheap to check and they explain sudden total drops far more often than link signals do.
- Open the disavow file and read every line. Mueller's advice was to make it what you actually want it to be. If nobody at the company can explain why a given domain is in there, that is a reason to question the line, not to keep it.
- Look at what else changed in the same fortnight: a migration, a template change, a redesign that dropped internal links, a CMS update. A structured technical audit surfaces these faster than staring at a backlink report does.
One more point on process. Removing a domain from a disavow file is not instant. Google has to recrawl those links before anything changes, so undoing a bad upload takes as long as the crawl takes, which is another argument for a small file or none.
What the report did not say
Several things people will want to read into this story are not in the source. Mueller did not confirm that the poster's ranking drop was caused by the disavow file, and he did not confirm it was caused by a spam update either. He said a spam update might have happened around that time and that he did not recall the date. He did not name the update, give a percentage, or describe any recovery pattern beyond saying that effects can take many months to appear.
He also gave no threshold. Nothing in the report says how many links is too many to disavow, which metrics justify a disavow, or how the tool works internally. Search Engine Roundtable reported that Google may remove the tool at some point, as Bing has, but reported no timeline and no announcement. Anyone presenting this story as evidence that disavowing is back should be asked which sentence in the source says so.
What this means for marketers in Thailand
Paid guest posting is a normal part of how many Thai sites have built links, often bought through agencies, freelancers or marketplaces, and often without the buyer keeping a record of what was placed where. That is the situation the Reddit thread describes, so the answer applies directly: the site owner who bought the links is the one with something specific to disavow.
The more useful read for Thai site owners is the boundary Mueller drew. If a Thai-language site loses positions across the board, a short disavow file is not a sufficient explanation, and the effort is better spent on the two causes he raised. Bilingual sites add a category he did not discuss but which behaves like his technical example: hreflang and canonical errors between Thai and English versions can strip a whole language from results while the other one looks fine. Anyone running SEO across Thai and English should confirm which locale actually lost visibility before assuming a link problem.
There is also a documentation habit worth adopting. If a supplier places links, the site should hold the list. Without it, an owner who later needs to act has no way to separate the links they are responsible for from the ones that arrived on their own, and the guessing that follows is what produces oversized disavow files.
FAQ about the disavow tool and ranking drops
Does the Google disavow tool actually do anything?
Yes. Google's John Mueller said on Reddit, as reported by Search Engine Roundtable on 2 September 2026, that the tool "does have effects," while adding that "it has its place & time." He made the comment to a site owner whose situation involved buying guest posts and links, and Search Engine Roundtable noted the statement was specific to that case.
Can disavowing a few links make a site drop out of Google?
Mueller said it cannot: "disavowing a handful of links is not going to make your site drop out of search." A large drop that follows a small upload therefore needs another explanation, and Mueller pointed to a broader spam update or big technical problems as the places to look.
My rankings fell the week I uploaded a disavow file. Was the file the cause?
The source does not let anyone answer that for a specific site, and Mueller did not answer it for the poster either. He asked for a date, said a spam update might have occurred around the same time, and told the owner to rule out technical failures on the scale of blocking the whole site before settling on the file as the cause.
Should my site have a disavow file at all?
Mueller's instruction was to make the file what you actually need it to be, "which might be empty, or might not be empty." Search Engine Roundtable added that most of the time sites should not disavow, that Google has said the tool hurts many more sites than it helps, and that Bing has already removed its version.
How long does recovery take if a spam update caused the drop?
Mueller said fixing issues around bigger algorithmic changes "can take quite some time - sometimes many months" before the effects show. He gave no specific timeline and no recovery percentages, so any more precise number being quoted is not from this report.
Where to go from here
The practical takeaway from the 2 September 2026 report is a change of order. Date the drop first, separate sitewide losses from page-level ones, rule out the technical failures, and only then open the disavow file, with the aim of making it honest rather than making it longer. If a site's link history is undocumented and a drop is unexplained, an audit that establishes what actually happened is a better first purchase than another round of uploads.
The original report is at Search Engine Roundtable. If a site's link history is undocumented, the data already sitting in Search Console is usually the cheapest place to start.
