Archives for December 2007

Selling links that pass PageRank

I’ve talked quite a lot before about buying or selling links that pass PageRank. Today I wanted to walk through a concrete example of paid posts and show why the major search engines don’t want to be affected by links within paid posts. The problem is best illustrated by a serious example: suppose you just visited your doctor and got a scary surprise: you or a member of your family have a tumor. The doctor is throwing around words like steroids, surgery, chemo, and radiosurgery. Much of what the doctor says washes over you, but you remember the word “radiosurgery” and resolve to find out more when you get home.

At home, you fire up a search engine and type “radiosurgery.” If you go to Google and search for [radiosurgery], the results give a pretty good overview of what radiosurgery is (focused radiation that targets tumors). In the first several results, you have balanced information from the International RadioSurgery Association, an overview page on Wikipedia, a great background page from RadiologyInfo, even an introductory-level essay on radiosurgery from the Mayo Clinic. If you came back from the doctor and wanted to get an overview of radiosurgery, I hope you’d find the results useful.

Now, think about how you would feel if your medical search was influenced by pages like this:

Screenshot 1

The posts themselves don’t mention it, but entries like these often turn out to be what are known as “paid posts.” That is, someone paid money in order to receive a review, and the paid review includes a link with the word “radiosurgery,” for example. There’s no disclosure inside these entries whether these posts are paid, nor do the posts use the nofollow attribute or some other mechanism so that search engines aren’t affected.

Now I’m going to ask you to put on your regular user hat. If you’ve just learned that you or a family member have a tumor, would you prefer that radiosurgery overview article from the Mayo Clinic, or from a site which appears to be promoting a specific manufacturer of medical equipment via paid posts? My guess is that you’d prefer the Mayo Clinic.

In the example above, notice that the post says “I myself had never heard of this treatment process until now.” On a subject as serious as brain tumors, it’s troubling if someone is getting paid to review and link to a site, especially if it appears that they weren’t aware of this treatment until they were paid to write about the subject. In researching this incident, we saw lots of people doing paid posts about brain tumors who admitted that they weren’t familiar with the subject beforehand. Sometimes posts were even more inaccurate:

Screenshot 2

“Is there any new medical breakthrough in the treatment of brain tumore [sic]?” If the paid poster had researched their subject more, they would have discovered that this particular treatment has been around for two decades. In fact, the promoted site mentions that one of the main reasons to use their technique is because it’s well-established, not brand-new. If someone doesn’t do enough research for a paid post to know that a treatment is well-established instead of brand-new, how can you trust their opinion about brain tumors? Here’s another post that claims this treatment technique is new:

Screenshot 3

“happy and proud to introduce the GAMMA KNIFE”? Again, this is not a new treatment technique. In fact, anyone who read the first page on the promoted site would have realized that one of the selling points of the device is that the technology is well-established. But there are more noticeable mistakes than calling the technique new when it’s actually been around for decades. Sometimes the posts don’t even get the name of the treatment right:

Screenshot 4

Notice that this post consistently refers to the “Lesksell Gamma Knife” treatment. The correct name is Leksell. I also saw one person call it the “Gama Knife” treatment, and someone else called it the “Gamma Knive” treatment. If you’re getting paid to write a review, shouldn’t you at least perform the basic research to get the product name right?

I have a lot more snapshots I could show, but I hope these examples help explain my point. For this very important (potentially even life-or-death) medical topic, we saw paid reviewers admit that they knew nothing about a treatment before getting paid to post about it, or who didn’t research the subject enough to know that a treatment was decades old instead of brand-new. We saw people writing about brain tumors who didn’t even spell “tumor” correctly, and we saw people who got the name of the sponsor wrong.

If you put your user hat back on, I hope you’ll agree that you wouldn’t want a serious medical search for brain tumor treatments to be affected by inaccurate or uninformed posts. In fact, if you stumbled across these entries on the web, you might not know whether someone got paid for writing these posts. In the same way that a regular surfer would want disclosure to know if a post were paid, all the major search engines also want to make sure that paid posts are adequately disclosed to search engines as well. Google’s documentation for webmasters gives examples of how to do that. I believe the vast majority of our users don’t want our organic search results for something as serious as brain tumors to be affected by links in paid posts.

I hope these examples help to explain the motivation for our quality guidelines, and how those guidelines ensure a better experience for users. To read more about this subject, you can start at Google’s quality guidelines. Where you see the guideline “Don’t participate in link schemes designed to increase your site’s ranking or PageRank” you can click on “participate in link schemes” to get more guidance. That page specifically mentions “Buying or selling links that pass PageRank” as being against our quality guidelines, and it’s also clickable so that you can read more about buying or selling links in our HTML documentation.

(Just as a side-note: I don’t normally send my blog posts to Google’s public relations or legal folks before posting. For this entry, I did get this post approved by Google’s PR/legal department first to make sure that what I said was an accurate reflection of Google’s opinion on this subject. If you’re concerned because this is a post on my personal blog, we also did an official blog post about this issue today.)