Personalization has been creating a lot of hype in the Utah search engine optimization scene lately. Personalization has been implemented into the results of Google accounts that are signed in for quite some time now. Google collects your search history and displays results based off of what youíve searched for and clicked on in the past. Many search engine marketers and optimizers have been reluctant to Google pushing this mainstream. Reason being is that it makes the game of SEO a lot harder when everyone has a different set of results.
Matt Cutts, Googleís spam king, claims that it will be tough times for black hat SEO artists because SEOs will have to focus more on users and less on algorithm reverse engineering. I was curious to what black hats were thinking so I cruised over to SEO Black Hat and found this post. Apparently, they are excited about personalized search and view it as a great thing. Additionally, nobody posted any comments about how this might affect their jobs as blackhat SEOs.
I donít think thereís going to be too much to worry about for SEOs. Google Bookmarks are going to have a huge impact on peopleís results. With that said, I see social bookmarking campaigns being just as big, if not bigger, than link building campaigns.
Marissa Mayer, Vice President of Search Products and User Experience at Google, said that personalized results only show up in about one of every five searches, and would only lift two results into the top 10, never replacing the number one organic result.
There are many caveats I want to address with personalization. The top result is typically the most visited result for any search query. If the top result never changes then thatís not really much personalization now is it? If youíre someone who typically uses the ďIím feeling luckyĒ button in Google then youíll always get the same result whether youíre signed into Google or not. Also, everything you click on gets ranked higher in your personalized results. How does Google determine if you liked what you clicked on or not? Thatís a major flaw in this system right now.
Another issue is the latency involved with temporary search patterns and localization. What if youíre on vacation for two weeks in New York and youíre searching for local restaurants, hotels, clubs, etc. When you return home your results are going to be skewed towards New York. How long will it take to recognize youíre not searching for New York based businesses anymore when you return home to Utah or wherever? What if your kids hop on your computer to search for something like kids games? The next time you search for games online your results will be tainted.
Personally, Iíd like to see Google more focused on semantic mapping and clustering of search terms while it refines personalization. They already have started implementing this feature for select search phrases, but I think they need to push it a bit more. I think Clusty.com and Quintura.com have both done a good job of semantic clustering.