Graph Search

What is Graph Search?

Facebook's technology for finding people and things that are socially connected

What is Graph Search For?

  • "Find people who share your interests"
  • "Explore your world through photos"
  • "Discover restaurants, music and more"

How Does Graph Search Work?

The Facebook search bar will suggest top search terms as you type. Search terms include all social objects : people, pages, places, apps, and groups to help users discover things that are most relevant to them.

The results page returns unique results tailored to the user who is using Graph Search. This provides incredibly relevant results to the searcher. Results that are socially applicable to the searcher will be displayed higher in the rankings. Also, brands that have been engaged with (checked-in, reviewed, liked, etc.) by the searcher's friends will be more prominent in the results.

Fun fact : When social results are minimal, Bing search results are displayed to supplement the search.

How Does Graph Search Rank
Search Results?

Facebook's goal is to display the most relevant search results at the top of the page. Facebook leverages a concept known as Term Frequency-Inverse Document Frequency (TF-IDF) to provide more interesting search results. Josh Constine from TechCrunch outlined a helpful use case:

Ask Facebook Graph Search for "books liked by founders" and you won't see The Da Vinci Code or The Bible first, though they're among the most Liked books. Instead you'll see books disproportionately Liked by founders compared to the general population thanks to TI-IDF. Surfacing The Tipping Point, and Design For Hackers shows how smart Graph Search is.

Facebook doesn't solely rely on TF-IDF to provide search results. They attempt to blend popularity, as well as TF-IDF, to provide context and value to the most popular objects. Using this blend of methods, along with exclusive social signals that Google lacks, could provide incredible results Google would be unable to replicate.

How Do People Use Graph Search?

People are searching predominantly for friends and photos. The third most searched category is Places. Users searching for Places provide businesses an opportunity to leverage Graph Search to their advantage.

Users have been querying Graph Search using slang search entries for the term "friends". They've been using slang such as "homies" and "peeps". Facebook has adjusted their algorithms to adapt to the unexpected slang, and Graph Search now returns correct results when using popular slang terms!

Below is the top search queries in a random sampling done by Facebook. You"ll notice that Places is ranked 10th overall. As mentioned above, Facebook has stated that the third most searched category has been Places. This list also reflects this:

  • 1 : Photos of [X]
  • 2 : My Friends
  • 3 : People named [X]
  • 4 : People named [X] who live in [Y]
  • 5 : Friends of [X]
  • 6 : Photos by [X]
  • 7 : Photos of my friends
  • 8 : Photos liked by me
  • 9 : Photos liked by [X]
  • 10 : Places near me

Can You Advertise With Graph Search?

Can You Advertise on Graph Search?

Facebook recently (April 16, 2013) began testing ads within Search Results. The early test ads were not targeted towards search queries, however we fully expect this to change.

Currently, the only option to advertise on Graph Search is through Sponsored Results. You can learn more about Sponsored Results from Facebook Help. Sponsored Results currently only display within the search bar auto-complete dropdown, but we believe these will eventually be displayed among the actual results.

We expect that brands will be able to use search data for Retargeted Ads in the future. We also anticipate relevant ads similar to Google Search Results.

Optimize Your Brand for Graph Search

1 : Drive Check-Ins and Reviews

Both are very important factors when it comes to being ranked within the search results. Facebook places more emphasis on these interactions to create affinity than measuring total "Likes" (fans). These interactions provide two unique aspects of data that a simple Like cannot provide, affinity and a sense of value (weight). Encourage your customers to Check-In and Review whenever applicable.

2 : Complete Your Facebook Page 100%

Any fields that are left open, should be filled in. This includes fields such as address, hours, and anything else that may be missing. Facebook will be using this information to index your Page. Without all possible input signals, Facebook might not serve up your Page in the results.

3 : Continue To Build a Fan Base

Many of the search results are ranked by total social connections. It's simple—the more fans, the more likely the brand will appear in the results. Warning : Don't resort to "buying fans" or using cheap tricks to attain fans. The quality of fan base is still important, especially when attempting to reach this audience organically. In the future, Facebook may examine engagement of particular posts to determine search results.

4 : Develop Engaging Content

Creating social connections between users is the key to receiving search results exposure. In the near future, we fully anticipate Facebook to incorporate Page's posts and fan's Comments. We also anticipate Facebook leveraging URLs that have been Liked in relation to the brand. This creates an incredible opportunity to drive people to your content, which can ultimately lead to more engagement.

5 : Add Open Graph Meta Data To Your Website

Open Graph Meta Data is essentially more tags in the <head> of your website. These tags are specific to sharing information on Facebook. Brands are able to tell Facebook how they want their content represented when shared. Control things like your Page Title, Description, and Images that are displayed.

Here's a sample of Open Graph Meta Data code :

<meta property="fb:app_id" content="YOUR_APP_ID" />
<meta property="og:type" content="goodfood" />
<meta property="og:title" content="Pizza" />
<meta property="og:image" content="http://www.example.com/pizza.png" />
<meta property="og:description" content="Hot Pizza!" />
<meta property="goodfood:pizza" content="http://www.example.com/yum"/>

Learn how to setup Open Graph Meta Data here

The Future of Graph Search

We'll be updating this site with information and tactics in the future, sign up below to stay up to date with Graph Search :