Many of us in SEO espouse the great benefits of Google Authorship and the potential of the rumored AuthorRank. Many in SEO would argue that AuthorRank has not yet been implemented as part of the Google algorithm. Or has it been added already?
A few weeks ago, one of my paid search clients contacted me with a problem. His normally very high ranking website had been dropping dramatically in traffic volumes from Google over the past 10-12 months. This particular client also is an association, a publisher of valuable information for the public.
After taking a look through Google Webmaster Tools, Google Analytics and other tools to determine what might be happening, it became abundantly clear that there were certain pages of the site (very popular pages) that were the main culprit in the traffic drop in Google. While (not provided) might normally make it difficult to ascertain which specific keywords these pages potentially lost ranking on, the keywords on these pages are very specific because they are pages for very specific health subject matter.
Suspecting a Panda update (or series of updates) might have been the cause of the rankings losses for these popular keywords, I took the copy from a few of these pages on the client’s site and checked Google for duplicates. Interestingly, the client’s site appeared in many cases to have been completely removed from the index, but other websites (with the exact same content) appeared.
This is not the first association I’ve had come to us with this problem. Associations often publish informative content for their members or the public, and sometimes, unfortunately, their copyrighted content is copied verbatim and used on other websites.
Sometimes this occurs when association members (such as doctors who are members of a medical association) seek to share valuable information with their own clients (such as their patients) on their own websites. Often, I think this misdirection is innocent in nature, but it does mean that associations (and other content publishers) face a daunting task of protecting their content online, especially if one of these other sites outranks their own site because of Panda updates.
Further Diagnosis: Authorship
While the Panda updates clearly seemed to be at play here, I still found it difficult to believe that Google would not still rank the site highly based on its incredibly authoritative inbound links from sites including the CDC, NIH, CNN and many state departments of health. Clearly, this site is an authoritative site based on inbound links from other authoritative sites as well.
As I began to dig further and take a closer look at the other sites that had been plagiarizing the client’s site content, I found another website that was not only using the content but also had implemented authorship on these pages. In total, this website was plagiarizing upwards of 89 pages of content, and each page was marked as authored by them. Could this have been part of the problem?
A Few More Relevant Details
Some other things to note about this problem include that the offending website is a locally-based business in Texas. As a searcher based in Virginia, you wouldn’t normally expect to see this local business high in SERPs based on geographic settings.
However, this site ranked very highly for very popular keyword terms, ranking alongside highly authoritative sites on the given keywords and subjects. The site had few, if any, inbound links. After doing some research using the Wayback Machine, it was also clear that these pages were likely added in the May 2013 timeframe, so they were relatively new pages.
Dealing with the Problem
After the client notified the offending website, the owner of that site immediately took the content down. Within days, the organic traffic to the pages from Google (and the rankings) also began to increase.
After only one week, we found that of the 89 pages that had been copied and claimed via authorship by the offending site, 60 percent had begun to rebound substantially in organic traffic from Google, some with as much as a 125% increase in traffic from Google organic. Below is just a sampling of our initial findings.
|Keyword ID||Google Organic Traffic Improvement in One Week|
You might theorize that this one offending site was the only example of duplicate content for these pages. However, in the case of page 1 from the table above, which has so far realized a nearly 24% improvement since the offending site was removed, there are still at least two other sites that also have the entire page duplicated.
In the case of other pages that saw increases, some had as many as 23 or more cases of duplicate pages and still saw significant increases (such as 15% or more).
As you would suspect, the traffic data indicated what we thought would happen… rankings began to recover after the offending site pages were removed. In the case of several keywords, we found that just removing the one offending site’s pages that were marked with authorship recovered rankings completely.
In one example that we benchmarked two weeks ago, the page for one keyword disappeared completely from Google’s index. Upon removing the offending site’s version with authorship, the page on the client’s site bounced back to a top 10 ranking in two weeks, even though there are still over 259 examples of exact duplicate content appearing within Google’s index from other sites.
While it’s still early in the treatment of the Panda issue for this client, by just cleaning up one site — one with authorship associated with it — the pages that had been losing traffic have clearly been back on the rise. So what does all this mean?
My theory is that Google is using authorship, even now, as a signal of ownership and origin. In other words, if authorship is applied to a page, my theory is that Google is using this markup as an indicator of the true owner of the content.
That’s a bit disturbing to think about because it would mean that publishers could have their content essentially hijacked by spammers who choose to apply authorship. And in many cases, as many SEOs have attested, an individual author may not be appropriate for certain content (as I would attest in this case).
This leaves publishers that refuse to implement authorship vulnerable not only to the regular onslaught from Panda because of others re-using the content on their own sites but also to losing visibility in the search engine altogether due to invalid authorship being applied.
I personally hope that Google isn’t giving this level of value to authorship. Like many things in search, the days of authorship are still like the Wild West, and many kinks and problems are clearly being worked on by Google. But granting carte blanche to various individuals who claim to be original authors isn’t a good way to determine authority.
Google, for its part, needs to be aware of when an original piece of content was first visible on a website (when it was first indexed) and give that version leverage over someone just claiming to be the content’s author on another site, because this may be creating potential for essentially rankings theft.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.