Also find me on...
Google Disavow Tool is a tool where you can submit a list of domains and urls linking to your site that you feel may be causing harm to your site’s rankings. Google will then discount those links when assessing the rankings for your website. So if your site rankings have improve significantly after the disavow request, it must be the successful effects of the disavow, right…?
But wait a minute…is that really so?
Recently, there is a post on Moz.com by Josh Bachynski discussing whether Google disavow is what it said it is or just a whole load of bulls**t. It has sparked quite a conversation (154 comments, last I checked) and even a respond on Search Engine Watch regarding Bachynski statement.
Bachynski published on Moz.com stating that it is inconclusive, based on his research, that the disavow tool recovers sites from Penguin penalties. It is likely due to other factors such as link attrition that may have fueled the recovery. He said:
The disavow file alone does not help you escape from Penguin, maybe in combination with deleting links, I’m not sure. I’m saying the disavow file is, unfortunately, the opiate of the masses. It is a safe myth we believe in because it makes us feel warm and snugly at night. But I’m afraid that, after scientific testing, I cannot prove that that is the case.
According to Search Engine Watch’s post, Marcela De Vivo chipped in and provided some evidence that the combination of link removal as well as a disavow submission would help websites penalized by Google’s Penguin Algorithm to recover.
She asked a very valid question which was also in my mind when I was reviewing Bachynski’s video post.”Is it possible that a site will not have any natural link attrition?” (more on natural link attrition later).
Bachynski built his hypothesis by the following:
Disavow Tool can be concluded to work if the site rankings improved only if NO links were lost during the period the disavow was made and the recovery. (NOTE: There is a lead time between the disavow request to recovery. This can range from several weeks to months, depending on when the next Penguin update is.)
So in other words, Bachynski would not attribute the reason for recovery to Google disavow if he noticed any links lost within the period of time.
He studied 12 websites that recovered from Penguin penalty just by disavowing the bad links. However, he was unconvinced as he noted that all 12 websites also had links lost over the period of time.
I feel that this hypothesis is brilliant. However, the key criteria (no lost links) is one that will be difficult to observe. It is highly likely that any website would have some natural link attrition, outside of the site owner’s control. Natural link attribution are caused by many factors such as a website where the link was placed is shut down, link being removed by webmaster of the site where the link is placed, Google deindexing the site where the link is placed, etc. All of these are outside the control of the webmaster whose website is previously receiving those links.
Hence, there is a need to dive deeper to analyze the quality of the links that were lost. Some of these links may be from high trust website which may in turn be passing valuable link juice to the website. Losing many good links from different sites will not help improve keyword rankings, it may negatively impact the site’s rankings instead.
Another factor to consider about lost links are whether the links are nofollow. If the links are nofollow, then it would not affect the rankings of the site even when they are lost. (Remember: Google disavow merely treat links on the domains and urls you’ve submitted as nofollow so that they don’t affect your rankings in future. Likewise, lost links that are nofollow also don’t affect your rankings.)
The hypothesis is logical, but it is very difficult to observe in reality. Most sites, if not all, will have natural link attrition over time. Regardless, it’s still a thought-provoking and controversial argument.
To investigate the subject deeper, I decided to make some observations myself too. Recently, I’ve also notice a rebound on keyword rankings for a penalized website.
But before that, let’s take a quick look at how Google Disavow works….(for the benefits of readers who are unfamiliar) 🙂
When You Should Use Google Disavow and How it Works
If your website has a manual penalty slapped on it or you’ve received a notice from Google in your Google Webmaster Tools saying that you have “unnatural inbound links”, you know that you need to clean up your site’s link profile by removing any harmful sites linking to you, before you can recover your website.
In a manual penalty, this requires you to do a few things:
1. Contact the webmasters of sites which you think are causing harm to remove the backlink to your site. (The webmaster may comply with your request or choose to ignore you. If they decide to ignore you, there’s nothing much you can do. Others may ask for a fee before they remove the backlink.)
2. Keep a record of the backlinks that you were able and unable to remove. Submit a reconsideration request to Google webspam team along with your record.
3. Disavow domains or urls where you could not remove the links.
In an algorithmic penalty slapped by Penguin, the process is a lot simpler:
1. Proceed to disavow sites that you think are causing harm to your site. (There’s no need to submit any reconsideration request to the webspam team).
That said, since you did not receive any warnings from Google, you have to first make sure that your site is indeed hit by a Penguin penalty.
There are free tools like Panguin Tool that will overlay your Google Analytics data with the Penguin update dates. So if you see a significant drop around the time of update, it is possible that your site is hit by Penguin penalty. I’m oversimplifying here as you also need to investigate your site codes and structure, content for duplicates etc, to make sure that your site is not affected by other issues.
Once you have submitted the Google Disavow file, the next thing and only thing you can do is WAIT. (You can however start to focus on earning relevant editorial links!)
It’s a 2-step process. First, it takes several weeks (according to this post) to have your disavow submission “registered”. Second, after it’s “registered”, you’ll still have to wait till the next Penguin update before you see any effects of your disavow. So if you have disavowed your links just a few days before a Penguin update, sorry friend, you have to wait for the next update.
On that note, it means you should not submit your disavow file on a regular basis because you may end up missing the boat. Each time you submit a new disavow file, it resets the lead time for step 1.
How A Disavowed Site Performed After Penguin 3.0
The website is www.dog-grooming-training.com. It is a site that promotes dog grooming to dog lovers. It belongs to Donald Lee and I’m publishing the site’s stats with his permission.
The site hasn’t been updated with any fresh content since Apr 2013. We redesigned the site in Jan 2014. In Feb 2014, I submitted a disavow request to discount the bad links, after rankings and organic traffic tanked in the aftermath of Penguin 2.0 update. We did not manually remove any links.
The site hasn’t been updated since then, so there are minimal changes that could otherwise affect the rankings of the site. In retrospect, the rankings would gradually drop over time since the site has been dormant for some time.
To conclude with sufficient confidence that Google Disavow did really help to recover the site, I would look at a few things:
1. Organic rankings for money keywords before and after the update. When a site is penalized, almost all of its rankings for keywords will drop and the number of keywords that the site is ranking for will also decrease significantly. If the site recovers from a penalty, we should see the rankings and number of rankings increase.
2. Organic traffic to site before and after. As a result of the drop in rankings and the number of keywords ranked, the organic traffic to the site is impacted. When the penalty is lifted, we should see an increase in keyword rankings and traffic.
3. Links lost over the period and the quality of those links. This require more analysis to determine if the links lost are good or bad links. Links lost that are bad links will help the site’s recovery. Good links lost, on the contrary, may reduce the rankings of sites, or at best, not have any effects on the rankings at all.
Organic Rankings and Traffic (Before and After)
Unfortunately, Donald did not set up any third-party keyword ranking software like Micrositemasters.com or Gshiftlabs.com. Google will only keep the keyword rankings for 90 days, not to mention that most of which will be classified as “(not provided)”. So we only had data from 11 Oct 2014 to 7 Jan 2015 from Google. This data will not be sufficient as we would also need to have about 90 days of ranking data prior to 17 Oct 2014 when Penguin 3.0 update was released.
I found a workaround for this using the Top landing pages data in Google Analytics. Since the webpages on the site had not been altered, the keywords that each of the top pages are ranking for would remain more or less the same. Without knowing what the keywords these top pages rank for, we can still have a very good idea of how well they rank for their keywords. There is a direct correlation between the organic traffic and organic rankings. The higher the organic rankings, the higher the organic traffic.
So using this, we can investigate the organic traffic for the top pages and see if there are any increase in traffic before and after the Penguin update.
Here’s the overall organic traffic for all pages
The organic traffic increased from 1,628 (3 months prior to Penguin update) to 2056 (3 months after Penguin update). Remember that the were no changes carried out on the site or any deliberate link building and social sharing during this time.
Here’s a snapshot of the top 5 pages and the accumulative organic traffic 3 months before and after the Penguin update:
You would see that the organic traffic increased after the update, suggesting that the keyword rankings associated with these have also likely increased. I’ve attached the PDF export of the data here so that you can look at the top 50 pages. Most of the pages had seen an increase in organic traffic. (Download here).
Links Lost and Quality of These Links
I use MajesticSEO’s data to investigate the lost links.
Here’s the historic index data for links lost:
I downloaded the data for the links to the website. Given my MajesticSEO’s subscription plan, I could only download 5000 lines of backlink data. However, it should be a sufficient sample to estimate the percentage of good links and bad links lost within the period 26 Feb 14 (date disavowed) to 17 Oct 14 (Penguin 3.0 release).
From the 5000 lines of data, I remove all the links and kept those that were lost between 26 Feb to 17 Oct 14 and that were not nofollow. That leaves a total of 623 lost links from 108 referring (unique) domains.
Then I counted the number of links that had a trust flow of less than 10 (these will be considered as bad links). Here’s a summary of the data in table form:
So the site lost a lot of bad links during that period of time. These links were lost due to natural link attrition.
Summary of Findings
I have hoped that the analysis of this site would provide some solid evidence that the site recovered as a result of the Google Disavow submission. However, since there were a lot of bad links lost during the period, it is hard to conclude whether the site recovered due to bad links lost or Google Disavow.
It would be great if I can have another site that is similar to dog-grooming-training.com, that had not been updated, and had not submitted any Google Disavow. If so, we would be able to compare and see if the site actually did any better after the Penguin update. If a similar site without disavow recovers from penalty, then we could safely conclude that Google Disavow is not necessary for recovery, so long the bad links are removed (either naturally or deliberately).
Till then, I would say that it’s best to do both, remove bad links and submit a disavow request.
To find out the bad links and disavow them is not that difficult. I will show you how to do it in my subsequent tutorials so you don’t need to hire a firm to do it (if you ever need this). 🙂
I hope I’ve provided some insights about Google Disavow. Now it’s your turn!
What are your views of Google Disavow? Should we use it?