Facebook Competes with Google: Graph Search

Continuing a trend of using members’ profile information for such purposes as targeted advertising, Facebook is now taking those analytics and applying them to a user-friendly new tool: targeted searches. It’s known as Graph Search.

Yesterday, Facebook revealed that it’s now in beta testing for an internal search option that allows users to receive results according to the preferences of their friends and family members. For example, it will give you results based on what your “friends” have “liked” — and those results will then be ranked according to the opinions of your closest confidantes first, which is gauged by how much you’ve interacted with them. Results are also ranked according to how many of your top friends collectively have “liked” certain pages, places, photos and similar content.

The Wall Street Journal quotes these as sample search phrases:

  • “Music that people who like Mitt Romney like.”
  • “Movies my friends in San Francisco like.”
  • “Photos my friends took in the 1990s.”
  • “Friends of friends who are men and single in Palo Alto.”
  • “Languages my friends from college speak.”

That not only means website results, but also photos and business pages. For marketers, this news is important because it would mean that if a business page has a high amount of “likes” within a certain community, then when anyone in that community searches for related content, the business page would be boosted to the top of the search results list on Facebook. Ultimately, it’s an algorithm that rewards popularity and loyalty, which is perfect for social media marketing.

The remaining results will come organically from Microsoft’s search engine, Bing! This could be a bold move for Bing! to come into closer competition with top-ranking search engine, Google. Google has similarly tried to incorporate search engine results with social media through its Google+ services.

You can sign up for the wait list to receive a Graph Search beta invite starting this week.


Recent Posts