Back in May, Matt Cutts talked about how Google was looking to use additional quality signals to help certain authority sites rank higher in a particular sector, and that they were going to also add authority as a factor into the Panda algorithm to soften the effect of Panda for quality sites that were on the borderline of being affected.
When the Panda update was rolled out in July Google gave the following statement:
In the last few days we’ve been pushing out a new Panda update that incorporates new signals so it can be more finely targeted.
The video below (from around 4.30 mins) is where this was first talked about in May.
In mid-July the impact of this new more finely targeted Panda algorithm was felt with lots of authority websites seeing good recoveries. The data below from Searchmetrics shows a few sites picked at random from the list of the biggest winners from the week ending 18th July. These are all good quality sites that have lots of authority both in terms of user data and links.
I’ve looked through quite a few other sites that are known to be affected by Panda and most have not recovered in this manner. Voucher sites in particular showed no sign of any uptick in visibility.
So what happened?
Most of the sites positively affected by this that I could find are in the job or car classified market. This indicates that Google is getting better at understanding user data for sites that suffer from poor “return to search” metrics because of the industry they are in rather than because the site is poor. Nobody finds a job or a car from the first few sites they look at.
We have access to data which indicates that the recoveries are across a variety of keywords and a variety of pages with no discernible pattern which leads us to the conclusion the recovery is pretty much sitewide rather than just across higher quality pages. Speculation was that Google would turn Panda from a sitewide algorithm to a page by page one but we have no evidence to support this.
Of course it might be that Google is now better able to ignore the really low quality pages on a site and stop them from affecting the overall domain but this would be hard to analyse as low quality pages on a Panda hit site would not be getting any visits before or after July.
In summary, Google appears to have found some kind of authority metric that allows them to dial down the effects of Panda for sites that are authority websites in their sector but perhaps have been hit by Panda in the past because their return to search metrics are poor or maybe they had lots of low quality pages dragging down the good ones.
The key takeaway is that recovering from Panda now has a second avenue along with the usual content improvements – sites can now focus on authority and brand as well as content and hope that Google takes notice.
I’m interested in any other winners and losers from mid July if you would like to leave a comment.
