As Thanksgiving approaches, there are a few additional things for me to be thankful for this year. And as you can tell by the title of this post, they have to do with cute black and white animals who like to ravage websites. :)
In August I wrote a case study about a
very interesting Penguin recovery. A client of mine, who was hit by both Panda and Penguin during the April 2012 algorithm sandwich, finally ended up recovering from Penguin,
but during a Panda update. That struck me as odd, although I wasn’t necessarily complaining. I simply had some questions for Google regarding the timing of the recovery. For example, was there a Penguin update during the Panda update? Are Panda and Penguin connected somehow? And does Penguin bubble up to Panda? I never received answers to my questions, but I was happy that my client recovered. So I moved on.
Well, here we are three months later, and it happened again. I have another client that was hammered by Penguin on April 24th, losing 75% of their Google organic search traffic overnight. They contacted me in July to try and right the ship, and I started helping them in early August. In this post, I’ll explain more about how I helped them, their progress over time, and how they eventually recovered from Penguin during a Panda update. You can see their initial drop in traffic below.
On 4/24, They Experienced The Classic Penguin Drop in Traffic:
Two Important Points As You Read This Case Study – Panda and The Disavow ToolOn November 5th, my client recovered from Penguin. That was incredible news, but if you’re checking the dates of algorithm updates in your head right now, you’ll realize that November 5th wasn’t the latest Penguin update. It was the latest Panda update. Yes, that means my client recovered from Penguin during a Panda update (and this was the second time I had seen that happen). So once again, I had questions about the connection between Panda and Penguin.
But this case study isn’t just about Penguin recoveries during Panda updates. Since the disavow tool was
released in October, we were able to supplement our link removal efforts by using Google’s new tool. And I’m glad it was released. More on that soon.
Below, I’m going to cover my client’s recovery story in detail. I believe there are some important points that can help webmasters who have been hit by Penguin, but who may still be struggling to recover. Also, it’s important to note that my client was nice enough to let me write this blog post documenting their situation. They definitely understand that their own experiences can help others impacted by Penguin. I’m thankful for that too. Let’s begin.
When reviewing a site hit by Penguin, it usually doesn’t take long to find a troubling link profile. You typically find a high percentage of exact match anchor text leading to a site from a number of spammy websites, directories, link networks, etc. In extreme cases, the entire link profile is filled with problematic links…
Realizing the Potential Loss of Link PowerMy client’s link profile was in bad shape. Unfortunately, some of their previous linkbuilding tactics led to a few thousand spammy links. After heavily analyzing their inbound links, I realized that about 70% of their links were categorized as unnatural. In case you’re wondering, I used a number of tools to identify and analyze their links, including Open Site Explorer, Majestic SEO Tools, and of course, Google Webmaster Tools.
The Percentage of Unnatural Links That Needed to be Addressed:
Once the link analysis was completed, I knew that a few thousand links would need to be removed in order for my client to recover. And I also knew that the spammy links comprised a large percentage of their total links. That meant they may not have much search power when, and if, they recovered. Yes, we had a lot of work to do.
Matt Cutts and the Disavow ToolIt’s important to note that the disavow tool was not released at this point. I knew it was coming soon (based on Matt Cutts explaining this in early June), and I explained this to my client. Regardless, I wanted to tackle as many link removals as we could manually. In addition, we had no idea when the tool was actually coming. It could have launched within weeks, or it could have been months. We just didn’t know.
Excel Spreadsheet MadnessI love Excel. It can be an SEO’s best friend. But for Penguin, you need to be prepared for Excel madness, working with many spreadsheets and worksheets, with thousands of URL’s to analyze. Yes, it can get monotonous. And that’s exactly what happened in this case.
With thousands of links to check, I implemented a multi-phase approach. I wanted several sets of eyeballs on the links, including key people at my client’s company. What I might see as potentially spammy, they might see as an older and valuable link (based on knowledge of their own industry). I’m aggressive with my Penguin approach, but I don’t like collateral damage. And that’s especially the case when my client didn’t have the luxury of many quality links. I didn’t want them to lose any “good” links.
Our final spreadsheets started to take form. We had links flagged for removal, links flagged as possible removals, and then links flagged as safe. All spreadsheets were sent to my client for final review. Once we had the green light from my client, we were ready to rock and roll with link removal requests.
Manual Link RemovalsNow that our spreadsheets were ready, and spammy links flagged, I helped my client craft an email for link removal requests. By the way, I’m a firm believer that the company hit by Penguin should send the link removal requests, and not the consultant or agency. I feel it’s the proper way to go. So, my client moved fast and sent link removal requests out to many webmasters. Now we just needed to hear back from them and hopefully see some movement with link takedowns.
Some webmasters responded and took the links down. Some questioned why we wanted the links taken down, and my guess is their reaction was based on receiving other requests too. And of course, some webmasters just didn’t respond at all. I was completely expecting this, but hoped we could impact a good amount of unnatural links with this first wave of requests.
Phase I Results: Our initial link removal requests took down ~20-25% of the spammy links. Not bad, but not good enough. I knew we needed to have more links removed.
Phase II – Hello WhoIs DataThe second phase involved tracking down any information we could about the domain owners who didn’t respond. Note, many of these domains were spammy directories, so we knew we wouldn’t track down every person in control of the websites.
A modified spreadsheet was created with contact information we could find via whois data. Then my client contacted those domain owners via the newly-found information. This resulted in another 10-15% of the links being taken down. At this point, we had ~40% of the spammy links removed. To me, this was progress, but still wasn’t good enough.
During this process, we documented which links were removed, which domains were contacted, which ones responded, and which ones didn’t respond at all. We would need this data for the reconsideration request, which I cover next.
