Two weeks ago, Matt Cutts, head of the webspam team at Google, told search marketers that Google had changed the way it handles the nofollow attribute. This week, Cutts followed up that disclosure with a post on his blog, revealing that the change was made more than a year ago, and no one seems to have noticed.
How can that be true? Search engine optimization is all about testing. You'd think that SEOs, especially the advanced ones using a tactic like PageRank sculpting, would notice if one day the effects of their work suddenly and dramatically changed.
There are a few possible explanations for this, but first we need to step back and review what we're talking about here. As I wrote two weeks ago, the nofollow attribute began life as a tool to remove untrusted links from a site's link graph. Mainly, the idea was to stop blog comments from counting as editorial votes by a site owner, since the site owner had no control over them, other than to disallow links in comments or stop allowing comments altogether.
Later, Google advocated nofollow as a way to identify paid links, so Google would know that a link should not get the PageRank benefit of the linking site, and that the site owner was playing by Google's rules about selling paid links.
Most recently, nofollow morphed once again to become an advanced SEO tool for controlling the flow of PageRank in internal links around a site. This practice, known as PageRank sculpting, is applied by adding a nofollow attribute to links to pages that a site owner feels are less important to rank for, such as "about us" pages or privacy policies.
The common belief was that those nofollowed links were removed from the PageRank equation, so the remaining links on a page would divvy up the full amount of "PageRank points" available. As it turns out, that's the way it worked until about a year ago, according to Cutts, when Google changed its algorithm.
Since then, he says, Google still counts the nofollowed links when dividing up the total PageRank points of a page. To illustrate, if a page has five links, and none of them are nofollowed, each link would pass 1/5 of the page's PageRank points. If two of those five links are nofollowed, then what used to happen was that the remaining three links would each pass 1/3 of the page's available PageRank points.
Now, according to Cutts, if two of the five links are nofollowed, the remaining three each pass just 1/5 of the total PageRank points, since the nofollowed links still gather the points, even though they can't use them.
That would seem to be a pretty big shift in the way links are being handled. Yet, apparently, no one noticed when it happened. Or did they?
As I said, there are a few explanations for what's really going on here.
Scenario 1: Matt Cutts is Telling the Truth
I like Matt, and I don't think there's a devious bone in his body. I'm inclined to believe that Matt is telling us the truth, the whole truth, and nothing but the truth. It's possible that the change didn't have that much of an effect after all. Why would that be?
- Maybe PageRank sculpting itself was not doing as much as people thought.
- Maybe it was, but there are too many variables at play with the sites that were using it, so the change in Google's algorithm went unnoticed among other shifts.
- Perhaps in the process of implementing the PageRank sculpting, the SEOs made other changes to the site's structure that did have an effect, which still worked even after Google's changed algorithm.
Scenario 2: The Competitive Advantage Theory
As I said, SEO is all about testing, especially for advanced SEOs working on complicated sites -- the very people who would be using PageRank sculpting to begin with. In this scenario, Cutts is still telling the truth, but it turns out that some SEOs already knew Google had made the change, but kept it to themselves.
It would be in the best interest of those who knew to keep that info to themselves, since it gave them a competitive advantage. If their competitors were running around trying to optimize their sites using a method that didn't work, they wouldn't have time to do things that really did help.
Scenario 3: The Conspiracy Theory
There are those out there who think Matt Cutts is not telling the truth. The conspiracy theory version of this situation says that PageRank sculpting works, and works so well that Google is uncomfortable with so many SEOs using it. They think that Google never changed its algorithm regarding nofollow, and that it still works fine, but Google is trying to dissuade SEOs from doing it.
Scenario 4: The Pity Theory
A final scenario goes like this: PageRank sculpting with nofollow never worked, and Google is finally letting everyone know they're wasting their time. While it was fun to watch SEOs debate the intricacies of a technique that took lots of time and energy, Google started to feel bad for sending them on a wild goose chase for so long.
I doubt this one's true, since there are several SEOs that say they've used the technique and have results to prove it. But they could be suffering from some of the same effects as in the other scenarios, where something else was causing the boost in rankings.
So here's a poll: What scenario seems most likely to you regarding Google's recent disclosures on how it handles the nofollow attribute? Reply in the SEW Forums.
a. Google changed its algorithm some time last year, and no one noticed.
b. Google changed its algorithm some time last year, and the SEOs who noticed were too smart to tell everyone else about it, to keep their competition off-balance.
c. Google never changed the way it handles nofollow, but doesn't like the way SEOs are using it to manipulate Google's rankings, so it cooked up this elaborate ruse to get SEOs to stop using it.
d. PageRank sculpting with nofollow never worked, and Google is finally letting everyone know they're wasting their time.