Penguin 2.1 is here, and from what Matt Cutts has quoted it is supposed to be quite a minor update with only 1% of search queries affected. If that were the reality then few webmasters would have noticed it but the reality seems to be entirely different.
From looking at the buzz on the web, especially in major forums, a lot of sites have seen a big impact and Barry Schwartz confirmed this here.
So what happened and, most importantly, what’s changed? The short answer, it seems, is ‘more of the same’ with the screw being turned on several key metrics to further demote sites that still have existing questionable link profiles. Here’s a breakdown of the type of links that seem to have been targeted specifically in the latest update:
Beyond this though our own data has thrown up several other critical findings around how Penguin, and recovery from it, is affected. Let’s look at those now.
1. You can only recover on the next Penguin update
This statement is commonly found on the web and it makes perfect sense since the Panda update follows a very similar pattern. If you are hit with Panda, you will see a direct drop, then once you have made the necessary changes to clean up the on page issues within your site, on the next update you will recover.
Now this was, and has, always been true for Panda but it has become a lot harder to track since the filter is now running very regularly now as part of the main algorithm. I can understand how people would think that Penguin runs in a similar fashion, but this is not the case from what we see.
Over the last few months, we have looked at a lot of new clients that have come to us after being been caught by Penguin and we have data to prove that it is possible to see recovery outside of the standard updates. Take a look at the client below as an example.
In this example we were approached because they had lost rankings. Once we looked into when the actual rankings drop occurred it was clear Penguin 2.0 had hit them. Just to confirm there was no message in Webmaster Tools indicating a manual penalty.
The task was clearly to work on link analysis and profile rebalancing but I am not going to discuss how to go about finding and discovering unnatural links as that is a post all of it’s own. Instead please take a look at our earlier guides on this:
The one additional statement I will make to the link analysis and removal process of Penguin is:
2. Make sure you physically remove the links as well as disavow them.
The second point to make from our data analysis concerns the effort of manual link removal and it impacted the approach we took to the client campaign as above.
While manual removal was initially discussed as a necessary action as part of any penalty recovery process Google’s John Mueller recently muddied the waters slightly by suggesting that companies should not pay for link removal, just disavow them.
Even though I respect the fact that this information has come from a credible source in Google, I cannot turn my back on our own data and the evidence we have gathered here that actually shows that physically removing links has a quicker effect.
In addition to this, if the unnatural links are causing problems with Google’s algorithm, in time other Search Engines will catch up and devalue or penalise these links also. If you simply submit these links to Google’s disavow tool, that won’t have any effect on other Search Engines. It is better to get rid of them completely.
Back to the client campaign discussed earlier and we embarked on a three months link removal campaign based on these findings, alongside content led link building to both get rid of the unnatural over optimised links and replace them with high quality branded links. We then started to see gradual recovery…
Checking the date of that uplift there is no mention of any Penguin refresh or update. This recovery happened slowly over time and is still continuing to do so now. This has happened now on several occasions and it has taught us the following:
I know there will be a few people that will say that they have seen huge recoveries on Penguin updates and there will undoubtedly be opportunity to do that but the more data we look at the more it suggests that refreshes are happening more regularly now and recovery happens at other times too.
So, what is the conclusion to our dig into Penguin 2.1? Here are a few key takeaways: