The year is almost done. I’ve noticed quite a number of “Best of 2008″ posts floating around the search marketing industry, along with the proverbial prediction posts: what does 2009 hold for SEM? The SEO department here at Apogee batted this question back and forth during one of our team meetings to formulate a list of answers. I’m not going to delve into the entire list right now, but instead, talk a bit about one item in particular from that list: the personalization of search and rankings.
Let’s first start with a definition of personalized search. I asked a few friends not barricaded within the ivory tower of search engine marketing if they noticed they could manipulate search results on Google when logged into one of their Google accounts, or if they noticed the “Customized based on recent search activity” text at the top right corner of their search results. The consensus answer? “No.” So, as long as there is an unaware class, I am of it. Basically, personalized search is providing search results to users based on search history, search query intent and user location, and in Google’s case, SearchWiki activity, among other signals.

Of course, increased personalized search on its own isn’t much of a bold prediction. Rather, predicting the degree to which it will infiltrate search marketing, and search engine rankings in particular, is how you grab headlines. Well, I’m not here to join the “Rankings R Dead” team or the “Personalization Won’t Matter Much” team - I know, the grey area is boring and oh so non-polarizing, but I’m not looking for true believers here. Looks like I’m not grasping at headlines.
So how will personalized search affect rankings and how will we utilize them in SEO?
Personalized search will have the most impact on those queries whose relevancy is disparately dependent on location – restaurants and the like – and queries that are not location specific will be affected minimally, but could be affected by search history.
It’s logical that users receive different search results for a query such as “thai restaurants” if one lives in Ridgecrest, CA and another lives in Bremerton, WA. Location should be one of the primary signals in determining personalized results for these queries. However, it’s illogical that users receive completely different search results for queries such as “linguistics” based on location, but logical that search history slightly effect those rankings.
If search history reveals the person searching for “linguistics” frequents websites with videos, perhaps tweaking the algorithms so that websites with videos receive a bit of a boost makes sense. Nonetheless, the overriding desire here is for relevancy. Personalization does not occur in a vacuum where other users do not exist. Looking to the collective, through linking patterns, helps provide greater relevancy than merely one user’s search patterns.
Also, I don’t think the changes will be drastic for those queries where search history can play a role in determining rankings, because Yahoo! is going to anonymize user data after 90 days, and I think Google and Microsoft will not only set a similar limit, but will be forced to do so eventually. If this plays out as predicted, search results will only be personalized to a certain extent.
Finally, how will smart SEOs utilize rankings as a metric for success?
The first example, location dependent queries, would generally fall into the local search realm. I don’t foresee a major change in this department except that local businesses will be required to seek out SEO services from their hometown agency. Rankings are not dead. They are alive, well, breathing, eating, sleeping and working for you.
For those queries potentially affected by search history, I don’t think the changes will be drastic. It seems illogical to personalize search results for the sake of personalization while detracting from relevancy. Wikipedia works so well because it looks to the collective – groups are much wiser than individuals on their own.
Will personalization force smart SEOs to adapt?
Of course. Will it potentially make tracking certain rankings harder? Yes, but harder does not mean impossible. However, proclaiming rankings dead is akin to saying search engines are dead. Search is a cornerstone of the Internet. So long as people need to find information about things they do not know, rankings will matter. So long as the search results are organized (ranked), rankings will matter because they bring exposure, traffic, leads, sales and revenue. Rankings should have always been looked at as a means to these ends.