Matt Cutts, Google’s head of search spam, announced today that there will be an algorithm update targeting low quality exact match domains. He also noted that this update is not related to Panda/Penguin. I can only assume we will see some shifts in the search rankings this weekend, but Matt’s assured everyone this should only affect 0.6% of the the English/US search queries out there. A small number, but this update will surely affect a handful of keywords I watch to try and keep a pulse on certain niches.
This is a welcome update over here. There is nothing more infuriating then having all your hard work beaten by someone burning through domains like exactkeywordmatch-aaa.com, exactkeywordmatch-aab.com, etc. Now only if Google will stop ranking irrelevant websites that are cloaking based on user-agent, like they did many years ago.
Update (10-1-2012): Some reports show that this update happened over the weekend. Dr. Pete over at SEOMoz suggests that a small sample had been affected. While the data may indicate that’s the case, it is still not possible to prove that the rank dropping in certain sites was directly caused by this algo. update. Consider this fact. One of my favorite niches to monitor Google’s web spam updates is the ‘payday loans’ search. Affiliate marketers in this space pull no punches here. When I started to learn what actually works here, I started using these searches as my first test when Google rolls out an update. You will see me frequently go back to this search to test the effectiveness of an algo. update. This small sample solidifies at least one statement Matt made: this should only affect 0.6% of the English/US search queries.
The above screen shot is page one of today’s search results (minus the local results). The only legitimate sites in there seem to be the ftc.gov link and EEGInfo. But wait a minute, when you visit EEGInfo’s website, it’s about Neurotherapy? How is that related at all to payday loans? Are they looking for extra funding on the quick? I must know more.
I was already in Firefox, so the first to do is switch your user-agent. I would like to add here, I found it necessary to switch my user-agent to GoogleBot for many years, but old tricks somehow seem to be working again. And whaddaya know! The HTML title and meta description changed on the site. How could that possibly help though? The content and site structure didn’t change at all, what possible benefit could this be to a black hat?
Well, I had clicked on the search result in Firefox. I tried it in IE and Chrome next. Got completely different results this time. In those two browsers, without switching my user-agent, I got this page:
So ‘tactics’ that were supposedly obsolete years ago suddenly work again. I see this website over and over again in my spam tracking. They burn through domains on a monthly if not weekly basis. But the site design & template never changes. A theory of mine is this type of stuff starting popping up when Google Instant came onto the scene. They were so focused on getting the results to the user so quickly, that flaws in the algo. started getting exposed. Or some of their rules just cannot be processed that quickly. I mean, Google is trying to index all knowledge on Earth. Massive amounts of data we’re talking about here.
That’s why we get Panda/Penguin type refreshes, and that’s why affiliate marketers who run sites like the screenshot above burn through so many domains. Panda and Penguin are not real time (yet), so whenever those jobs are run against Google’s indexed data, a lot of these sites disappear from the search index. Then, in a matter of days, the exact same sites come back, just under different domain names, Penguin update happens again, rinse and repeat. That’s why I think the Google out ran it’s kick coverage with Google Instant, and we get spam clean ups every 6-8 weeks (guesstimating) the timeline). And now, they are trying to work their way back to having cleaner results.