How should you evaluate the results of disavowing backlinks flagged as “toxic” by a tool?

ProcrustesPR

A rise or fall in traffic after disavowing links is not enough to determine whether the action succeeded or failed. A tool’s “toxic” label is not conclusive evidence that a link is harmful. To evaluate the results, compare the same page’s performance over equal-length periods before and after the action, then consider its existing trends and changes reported by different data sources. Ahrefs’ 2024 experiment involving three pages found no clear overall improvement in rankings or visibility. This does not mean that disavowing links never works, nor does it establish that the disavowal caused a traffic decline.

ProcrustesPRProcrustesPRHow should I divide my DIY SEO budget?

Establish a reason for disavowing links—do not treat a score as a verdict

In an article dated 2020, Semrush cautioned that a high toxicity score alone was not sufficient grounds for disavowing a link, and that many flagged links were actually harmless. The article recommended checking for clear evidence of link buying, exchanges, or manipulation rather than disavowing links in bulk based on scores.

Ahrefs’ 2024 article also distinguished tool-generated flags from specific circumstances that warrant further review. It quoted Marie Haynes as recommending a full audit before submitting a disavowal file if a site had a manual action for unnatural links or a large number of links considered manipulative. This was professional advice offered at the time and should not automatically be treated as current official policy.

Compare equal-length periods and existing trends

On July 26, 2024, Joshua Hardwick disavowed 129 URLs that a tool had classified as “toxic” or “potentially toxic.” These URLs linked to three articles, whose performance he compared over the 20 days before and after the action. Total traffic reported by GSC fell by 7.1%, while Ahrefs’ estimated visibility declined slightly both before and after the action.

Interpreting these datasets also requires understanding their limitations. Patrick Stox said in the article that he would prioritize Ahrefs’ data, calculated using average search volumes, when assessing rankings and visibility. However, those data could update more slowly than GSC. Traffic reported by GSC, meanwhile, could be affected by seasonality and random fluctuations. Do not draw conclusions by selecting only the metrics that match expectations.

Examine each page to put traffic changes in context

  • Traffic to the first page fell from 574 to 505, but estimated traffic shifted from a slight decline to a flat trend. The author considered a small positive effect from disavowing links possible, but thought it more likely that a long-term decline had simply leveled off naturally.
  • Traffic to the second page fell from 291 to 267. Both GSC traffic and estimated traffic shifted from rising to falling, with estimated traffic dropping noticeably about ten days after the action. The author therefore suspected a negative effect from disavowing links, though this remained his interpretation of this particular case.
  • Traffic to the third page rose from 156 to 176. GSC traffic continued its existing upward trend, while estimated traffic declined slightly both before and after the action. The author judged the effect of disavowing links to be minimal or nonexistent; the growth could not be directly credited to it.

Twenty days was this experiment’s observation period, not a universal timeframe

The experiment was originally scheduled to run for a month but ended early because a core update began on August 15, 2024. The author considered nearly three weeks sufficient to observe results, but this does not establish how long every website should wait.

When evaluating results, record the action date, the before-and-after comparison periods, and whether those periods overlapped with a core update. If existing trends remain unchanged or metrics conflict, avoid attributing changes directly to the disavowal. The article also acknowledged that some SEO practitioners believe disavowing toxic links helps; this three-page experiment did not settle the disagreement.

ProcrustesPRProcrustesPRWhy is AI not mentioning my brand?

Disavowing tool-flagged backlinks: perspectives from across the web

Are links that a tool does not flag guaranteed not to harm a website?

No. Marie Haynes believes that SEO tools rarely identify the toxic links that can actually hurt a site’s performance in search algorithms. Tool-generated flags are therefore not a comprehensive guide to which links are harmful; the absence of a flag does not mean all risk has been ruled out. This is the assessment of the expert quoted in the article, not a finding about any individual website.

Can the earlier test that disavowed all links prove that disavowing only links flagged as toxic would produce the same result?

No. In a separate 2021 test involving the same three articles, Ahrefs disavowed all links, after which traffic fell sharply. Because the scope differed, that result cannot be applied directly to disavowing only tool-flagged links. However, the author drew a caution from it: if a tool mistakenly flags helpful links, disavowing them could harm traffic.

How should claims that disavowing toxic links helps be assessed?

Start by considering which links were disavowed and the website’s circumstances, rather than assuming every site will see the same result. The Ahrefs author acknowledged that some SEO practitioners disagree with his recommendation and believe disavowing toxic links helps. Based on his own data, however, he considered blindly acting on tool-generated flags unlikely to deliver a significant benefit and potentially harmful to traffic. Those differences in scope and circumstances still matter when assessing such claims.

ProcrustesPRDoes an AI brand mention bring website visitors?

References

  1. I Disavowed “Toxic Backlinks”: Here’s What Happenedahrefs.com · 2024-08-29T13:34:32+00:00 — Discusses: Are links that a tool does not flag guaranteed not to harm a website?; Can the earlier test that disavowed all links prove that disavowing only links flagged as toxic would produce the same result?
  2. How to Disavow Backlinks (+ Find Out if You Really Should)www.semrush.com · 2020-07-10T11:00:00Z — Source used in the article above.

Related articles

How should I divide my DIY SEO budget?Why is AI not mentioning my brand?Does an AI brand mention bring website visitors?How Long Does It Take to Evaluate SEO Results?How do I write titles customers want to click?How is organic traffic different from paid traffic?Why do I have keyword rankings but no customers?Are Google Maps rankings the same as website SEO?
ProcrustesPR
ProcrustesPR