In class, we learned about the basics of Google’s PageRank algorithm and how it assigns a ranking to web pages based on, amongst other factors, the number of pages that link to a certain page and the relative ‘importance’ of those links. With the advent of social media giants like Facebook, a new era of search that brings unprecedented personalization is in the works.
Facebook and Microsoft have teamed up to integrate Facebook data into the Bing search engine, in the hopes that Facebook’s social signals will allow Bing’s search results to be more accurate than Google’s use of link signals in PageRank. Bing Social, as it is called, will be a new section that combines results from Twitter and Facebook to deliver results in real time, using public profile data to detect trending social topics that could be relevant to the user.
Instead of a generalized search that only takes into account page links, the new Bing Social would integrate your network of friends’ preferences into its results. The information taken into consideration could be anything from a ‘like’ to shared photos, links, comments, location, status updates, etc. One example use of this information is more accurate people search. Based on people you know and interact with socially, this new search algorithm would be able to more accurately pinpoint exactly who you are searching for.
This has large implications for SEO (search engine optimization) strategy. Currently, one of the most important aspects for a web-based company is its placement on top search engines. Indeed, if a Facebook ‘like’ or a follower of a Facebook page becomes a strong signal relative to a link from an authority website, then companies would have to revamp their SEO strategy in order to maintain top search engine rankings.
There are many factors to consider when dealing with social signals versus link signals. How does one compare the weight of a Facebook ‘Like’ to a link from a webpage? Does the weight of ‘Likes’ and other shared material from Facebook depend upon who in your network shares them? And how would one measure the weight of a ‘close friend’ rather than just an acquaintance (surely one would value the opinion of a best friend over an acquaintance, or at least consider their inputs differently). Ultimately, these are all questions that could be considered in the implementation of the new search engine.