@softplus has stated a few times to ignore SEO tools toxic links/domains reports.

 

John replied, "The concept of toxic links is made up by SEO tools so that you pay them regularly." Update, John later edited this response to say, "Nothing has really changed here - you can continue to save yourself the effort."

 

<https://www.seroundtable.com/google-again-blasts-the-concept-of-toxic-links-37453.html>

When in doubt, leave disavow out.

Yeah, I am 100% in agreement with you. was hoping to get some validation.

 

Unrelated. Here is an actual pic of me when a CEO says they have SEO experience

That article has a few links of the same message, show that to the CEO and be like Bruh, we don't do that here

we don't do that

yeah, that's a good one. This one from Ahrefs does a good job too: <https://ahrefs.com/blog/toxic-backlinks/>

> we had them set a few of their child sites to nofollow because it was giving PBN vibes and we saw positive results.

@strager.andy do you think this was the cause? or coincidence with algo update / other explanation?

 

a thought: if there is no such thing as a 'bad' link, why not go ham and build as many possible links as you can, regardless of quality and relevance? bad ones are ignored, and good ones improve your rankings

We had done a ton on the site because they were getting hammered update after update. there was a big emphasis on UX which I think helped as well.

 

We ~disavowed~ (edit: nofollowed) about ten sites, and they were all pumping out ~20 articles a day each with multiple links per article to the main site. This had been going on for a decade. We saw significant improvements after the next core update. Hard to say for certain, but the timing sure lined up.

I think what is happening is that they saw improvements from disavowing and want to clean up the link index. Of course, there comes the whole educational process of disavow vs nofollow

Nofollow is a very useful option - at a previous company with 100m+ URLs to help reduce wasted crawling, I had the team nofollow a bunch of internal links which actually helped get the right URLs crawled/indexed.

Yeah, I have had a ton of success with it lately in publishing and ecomm. Everyone has been saying it, but for real, index quality is such a bigger deal than many make it out to be.

I think the aggregate of all "toxic link" reports cover 100% of a site's links. John's right, they're hilariously bad. My teams' SOPs have been "only deliberate schemes" ... if there was an obvious webspam footprint to something extra spammy that somebody did, since that's long been the spirit of how Google fights spam algorithmically. Or a clear negative SEO attack... super obvious when that happens, often 10K new links overnight with pornographic anchors.

 

I don't know what to think of the tool anymore, but we'll still use it narrowly for good measure. John's downplaying the value, but saying that it works the same as ever? Does he mean that it was always the equivalent of those disconnected "door close" buttons on some elevators? @cyrusshepard did an experiment where he disavowed all his links and nothing happened, but it's unclear if that's because the tool is bunk, or he hit some "measures to save you from yourself" that Google mentions often.

I think @corey brought up a very good point - SOP. *If you do use, I would suggest creating a written procedure for folks to follow.* It's easy for people to make mistakes.

 

And I'm in the camp that doesn't use them anymore.

I would only disavow if a client gets a manual penalty.

 

I had to do it with a client who was flagged after buying many optimized links. Anecdotally, Google also flagged some of the client's mini-sites that looked like PBNs, and we nofollow and edited the anchor text on these links (being overly cautious).

The semrush toxic links tool is pretty bad, especially outside US-EN. I've seen it flag legitimate editorial links from major local non-English publications.

Yeah, completely agree. My heart drops when I get an email about it. I am glad to see that is what everyone else thinks as well