CHAPTER EIGHT How Social Media and User Data Play a Role in Search Results and Rankings Search engines continually seek to improve the relevance of their search results. They do this by tuning the weight and mix of the types of ranking signals they currently employ, or by implementing new signals. Starting in 2010, evidence mounted to suggest an increasing weight on ranking signals from social media sources. In December 2010, Google and Bing both confirmed this in response to questions from Danny Sullivan (http://searchengineland.com/what -social-signals-do-google-bing-really-count-55389), and on January 10, 2012, Google announced Search, plus Your World, a revamping of search results that mapped in recommendations from your social connections directly into the search results. In addition, February 24, 2011, a long debate over whether search engines treated user engagement signals as a ranking factor came to an end with the release of Google's Panda update. While the algorithms will no doubt continue to evolve rapidly, adding new types of ranking signals to search engine algorithms is a tricky process that requires a tremendous amount of testing. As we discussed in Chapter 2, the Web consists of hundreds of billions of pages with fundamental differences in their construction and content. In addition, the needs these pages serve and the ways that users interact with them are equally varied. Search engines generally perform many levels of testing of new algorithms, including live tests with a small portion of their users prior to the general release of any new ranking factors. You can read more about all the stages of testing in this interview Eric Enge did with Google's Director of Research, Peter Norvig: http://www.stonetemple.com/search-algorithms-with-google 345