This week Mark Zuckerberg took the stage to introduce a new way for Facebook users to “discover content and information from their friends and the businesses and brands they’re connected to” called Graph Search. Graph Search is the third pillar of Facebook’s products. The other pillars being the News Feed and Timeline.
“Graph Search is not Web search. We’re not indexing the Web here. We’re indexing our map of the graph. Graph Search is very different,” stated Zuckerberg.
Graph Search is very different.
What is Graph Search?
Graph Search allows people to use their friends as filters for information in search results. For example, someone could use Search Graph for “sushi restaurants that my friends have visited in Cincinnati” or “movies my friends like.” Facebook then provides results based on a user's personal connections, potentially providing more contextually relevant search results because a person likely has a natural affinity to enjoy the same things or similar things as his or her friends.
Users can also find content that’s been buried by being on Facebook for several years. Searches for “Photos taken in Cincinnati” or “Photos of friends before 2000” will bring up content a user may have forgotten existed. They can also cross search for even deeper information like “Friends I work with who watch Game of Thrones.” Yes, the water cooler just got more interesting.
Facebook, of course, won’t always have the answer, and that’s where its partnership with Bing comes in. Bing will deliver Web results within Graph Search when relevant, which has the potential to take away a small amount of search traffic from Google. Google shouldn’t worry just yet, but if Graph Search does take off, it might have reason to.
As with anything Facebook does, this has the potential to raise privacy concerns, but Facebook has noted that search results are limited to what users are already able to see on Facebook. Content that is not public will not be shared, and only content users share with friends will show up in their search results, not the results of those they’re not connected to. Many users, however, are not clear on what they share, what they don’t and who they share with, so this has the potential to unearth content users hoped was buried forever. Users may discover their content wasn’t as private as they thought.
Why Does Graph Search Matter?
Graph Search is a version of search only Facebook could deliver. Google uses information such as website traffic, inbound links, social signals, recency and so on to deliver relevant search results based on keywords. And in most cases, Google does an incredible job.
Still, Facebook’s trove of user data and understanding of how users are connected allows Facebook to deliver results based on your friends. It’s not about analyzing the entire Web. It’s about analyzing users’ personal relationships, identifying intersections and delivering relevant and, potentially, more useful information based on phrases, not just keywords. Google understands this, which was a motivating factor behind the launch of Google+.
Facebook’s not competing with Google where Google is strong. They’re competing where they can win—with social data.
“Web search is designed to take any open ended query and give you links that might have answers," Zuckerberg explains. "Graph search is designed to take a precise query and give you an answer, not give you links that might provide the answer.”
More precise answers benefit users, and what benefits users, benefits Facebook.
Why Is Facebook Doing This?
Graph Search makes user data and the data of their friends more useful and more available to all. Facebook has always benefited from user data in the form of age, gender, relationship status, hometown, current city, likes, interests and so on for very specific ad targeting, but now, users can leverage more of the data of their friends to get relevant information they can act upon as well.
The potential benefits of Graph Search encourage users to do two things for Facebook.
- It encourages people to provide more data. A search for “friends who like the same books as me” is irrelevant if a user hasn’t updated his or her profile’s ‘Favorite Books’ section for three years.
- It gives users one more reason to stay within the Facebook ecosystem and not go elsewhere to find information.
These two actions can be repurposed and sold to advertisers.
How Do Marketers Respond?
At its core, Graph Search is searchable word of mouth. As users look for information like “restaurants my friends have gone to,” “my friends’ favorite movies,” “clothing brands my friends like,” and so on, they’ll find implicit third-party recommendations from the people they’re connected to.
There are currently no new ad formats specifically for Graph Search, but that doesn't mean marketers can’t take action.
Marketers should be asking themselves what they can do to show up in Graph Search results. Graph Search offers another way to reach more users, but the key is that reach will be relevant and likely come with an implicit third-party endorsement from a user’s social connection.
Here’s what marketers can start doing today…
- Optimize Facebook Pages: If your brand doesn't have a Facebook Page today, get one and start using it. All marketers should make sure the name, category, vanity URL and ‘About’ information are up-to-date as they will provide more information to help their businesses show up in relevant search results. Businesses with Place Pages should make sure addresses are up-to-date and even consider merging their Place Pages with their Business Pages if they haven’t already. All of this will create more information that Facebook can index and pull into search results when applicable.
- Encourage User Signals: Graph Search is built upon Facebook’s EdgeRank algorithm (an algorithm that works to make the most relevant content stand out for a user). Encouraging actions such as checking-in to a business’ location, ‘liking’ and commenting on content, ‘liking’ the Page and so on improves EdgeRank. Success in these areas starts with creating compelling content users want to act on and ensuring your offline experience supports your online one by, for example, adding in-store signage to encourage check-ins. Encouraging social actions will makes a business more relevant, which in turn means it will show up in more search results.
- Uncover Consumer Insights: Finally, Graph Search has the potential to be an excellent consumer research tool. Marketers can search for “TV shows liked by men between 18 and 26” or “brands liked by people who also like cooking.” This could pull out relevant connections that might not otherwise be made.
Where is Graph Search Going?
Facebook is full of data, and this is only the beginning. Graph Search is slowly rolling out in a limited beta. The initial focus is in four key areas: people, photos, places and interests. The long-term future will include the feature being available in all languages, on mobile devices and with the ability to search everything, including Open Graph actions and user posts.
The true power of Graph Search isn't what was announced this week. It’s in the long-game.
Gathering user search data allows Facebook to identify user intent. Intent has, for the most part, been owned by search engines. Facebook will soon be able to combine intent with where Facebook already excels, social context and hyper-targeting advertising. A movie studio could advertise the next Quentin Tarantino movie to the friends of people who like Tarantino’s Facebook Page who have searched in a way that indicates intent to watch or go see a movie. That is relevance.
The addition of intent could come to life on Facebook in a number of ways:
- Search ads, which already exist, will be ported over to Graph Search, but Graph Search will likely have its own ad unit in the near future that brings intent, targeting and social connections together for advertisers.
- Intent may also be implemented across Facebook’s other ad products, including the Facebook Exchange and its external ad network rumored to be in the works. Facebook could gather intent by looking at a search for “children’s toys” on Facebook and deliver an ad from Hasbro on a contextually relevant website using its network.
- Zuckerberg declared that Facebook is a “mobile company” last year, and one of the obvious shortcomings of Graph Search is that it’s desktop only. As the feature expands to mobile devices, businesses may be able to use Graph Search on mobile devices to deliver ads when users make queries like “restaurants my friends like that are nearby.” The feature would merge intent, social context and location.
But Let’s Not Get Ahead of Ourselves
Graph Search has a lot of potential, but it still has a lot to prove.
Facebook will have to train users to search on Facebook as well as take actions they may not already be taking: keeping their profiles updated, rating/reviewing businesses, checking-in at businesses and so on.
Privacy is another issue. Users who lock-down their profiles will only be supplying information to a certain segment of Facebook friends, making the feature less useful for everyone else. All it will take is a few privacy issues to make more people restrict access to their profiles and information. The feature is only as good as the data users provide and share. Facebook will need to encourage openness, while taking steps to protect user privacy. That’s a balance Facebook has struggled with in the past.
Finally, it’s important to remember Google+. Facebook has the same challenges with search that Google has had. The first challenge is getting users to understand the benefits of the feature and then get them to use it on an ongoing basis. This won’t necessarily be easy. Just as Google faced the challenge of getting users to think of it as being more than search but also social, Facebook faces the challenge of getting users to think of it as being more than social but also search. Google and Facebook are playing each other’s game. The first to master the other’s specialty will be on to something incredibly powerful.