The Great Disavow Debate
View original post via Search Engine Watch
A Whiteboard Friday video from Moz has stirred up significant debate in the SEO community. In this video, Josh Bachynski tackles the question:
Can uploading a disavow file to Google lift a Penguin penalty without removing links?
Bachynski states in the video that there are some SEO consultants that claim uploading a disavow file alone can get Penguin penalties lifted. In his analysis, however, he found that there is no evidence to prove that a Penguin penalty can be lifted with only a disavow file.
Link Attrition and Penguin
According to his analysis, the sites that had showed signs of recovery from a Penguin penalty had lost links in the prior months between the Penguin penalty and recovery, meaning the disavow file wasn’t responsible for the recovery. Since there was link attrition as well as a disavow file uploaded, we cannot isolate the disavow as the only factor. Therefore, the theory that a disavow file is solely responsible for the recovery doesn’t hold weight.
To make Bachynski’s claims valid, we need to consider the question:
Is it possible and realistic that a site doesn’t lose ANY links between Penguin refreshes?
Our research shows that this is not possible. All sites lose links for a variety of reasons, such as domains that were changed, not renewed, refreshed with new content, sold to new owners, etc.
If you study hundreds of sites using Ahrefs or Majestic, then it’s apparent that most sites lose links in the period in between Penguin refreshes. Even though Bachynski’s experiment is invalid due to link attrition, the question to be explored is whether you can recover with a disavow file with minimal link attrition or link removals.
Are Link Removals, Combined With a Disavow File, Necessary to Recover From a Penguin Penalty?
To explore this matter, we looked at a few client recoveries to correlate traffic recovery with link attrition. We correlate search engine traffic from SEMrush and link loss/acquisition from Ahrefs. You can see the graphs below:
By looking at the first chart for each site, and the lift in traffic that occurred around October, we can ascertain that these are true Penguin recoveries, and not recoveries from manual or Panda penalties.
In the second chart for each site, taken from Ahrefs, we can also see the pattern of link loss/gain.
In all of these recoveries, we see the following commonalities:
- A decrease in backlinks
- A decrease in referring domains
- An increase in referring domains after the decrease
Of course a few sites do not provide statistical significance, but when you look at the commonalities you can start seeing trends emerge.
The trend and commonality that we’ve seen in all of our Penguin recoveries is that they all lose a significant amount of backlinks/referring domains at some point before their recovery. Some build those links back after several months, while others don’t, but the link loss is consistent across domains.
Josh Bachynski’s analysis disproved the idea that uploading a disavow file alone is enough for a full recovery, but we see here that by combining disavows with link removals, the chances are greatly improved.
Of course, further studies should be done on hundreds of sites that have recovered, correlating their search traffic with the link charts on Ahrefs, to achieve statistical significance. Yet this does provide some hard evidence for commonalities between sites recovered from Penguin.