The Internet, Ampush Respond to Facebook Graph Search
January 16, 2013
Yesterday Facebook held a much-anticipated press conference to announce its latest product, Graph Search. Unveiled as a Facebook-specific tool to find answers to your queries, Graph Search delivers a set of people, places, photos, or other content that’s been shared on Facebook. Executives assured the press that privacy considerations are at the forefront of Graph Search, and it will never serve results that cannot already be found by browsing Facebook the old fashioned way.
Two former Google employees, software developer Lars Rasmussen and Tom Stocky, form the cornerstone of the Graph Search development team. While the product is currently in beta, users are encouraged to join a waitlist as Facebook will be slowly rolling out the feature to the community over the coming months.
Facebook hopes Graph Search will change the way people interact with the world’s most popular social website. The response from the press has been mixed, as some journalists such as Hamish McKenzie (PandoDaily), are inspired by the potential of Facebook’s latest tool:
….and now Graph Search helps unlock the latent potential in human relationships that otherwise might not have happened. Keep in mind, also, that, per Zuckerberg, Graph Search is only in Version One of beta. The capabilities Facebook showed off today are just the beginning.
And Meghan Kelly (VentureBeat), who had some time to spend with the tool, shared:
I really like the feature that lets you look at all the pictures you’ve previously commented on, or previously liked. It helps recall old memories, or locate photos you may want to disappear.
While others, such as Erin Griffith (PandoDaily), were decisively put off:
Hello, Graph Search. You are Facebook’s latest attempt to open up new revenue streams. Welcome to the Facebook income family, which now includes banner ads, in-stream ads, Facebook Gifts, and the quickly-shrinking revenue from Zynga games.
However, a Facebook spokesperson has said to the press:
We are not currently serving ads to people based on their searches. As before, ads can show up as sponsored results as part of the search suggestions when you type in the search bar. We are not launching any new ad capabilities at this time, and there are no ads in the search results page.
Based on how Facebook has behaved in the past, there is no doubt that this capability will be turned on at a later date. What this means for brands is that Likes, which have come under scrutiny by social media gurus and advertisers alike in recent years, will take on greater importance since they will be used as direct indicators for results in Graph Search as Business Insider explains:
Likes are now back on the frontburner at Facebook. When users search for restaurants favored by their friends, Graph Search will return results based on the number of Facebook Likes their friends have checked. The intent is to deliver search results that are useful because it’s like receiving an instant poll of your friends. Obviously, businesses with more Likes — more fans, in other words — will do better in this environment than those with fewer Likes.
Looking further into the future, Graph Search could develop into a much more robust and sophisticated algorithm. It may begin to include location-aware purchases (e.g, shopping or dining); referral purchases/leads (e.g., your friend John got a Honda Civic and he loves it); as well as tailored purchases/leads (e.g., you like the show ‘Homeland’ and live in a densely populated city, so you are 200% more likely to buy a Honda Civic). Obviously these things already exist in some form or the other (e.g., Hunch.com), but Facebook has the financial means and executive support to actually allocate computing and brain power to make these recommendations accurate and far more usable — not to mention the critical mass of users necessary to make it work at scale for advertisers.
Although Graph Search results don’t currently integrate Open Graph app data, there are few obvious implementations that could improve the offering right away. Some popular examples include Yelp (e.g., restaurant ratings), Spotify (e.g., suggesting music your friends like), and Foursquare (e.g., activity suggestions based on your friends’ Likes and check-ins). Having a robust search product like Graph Search should help further catalyze and accelerate Open Graph adoption as it provides a clear incentive that advertisers can grasp today.
What will also be interesting is the extent to which Graph Search campaigns are constrained by search volume which will be influenced by how well Facebook markets this product to users. Ampush has occasionally observed volume limitations when running sponsored results for some large-scale campaigns.
Still, Graph Search is huge for at least two reasons: intent and utility.
The first reason is particularly obvious: now Facebook is capturing user intent and are entering the search game. Users will initially start by looking for people, pictures, photos and places. But why stop there? Why not use it for movies and TV shows or a product someone is considering purchasing? Ads will come into place both directly and indirectly. Since Facebook can now harness intent (with their unique and proprietary data), advertisers will be very interested in placing a “sponsored result” (with social context) for a given query. Indirectly, Likes and other ‘social’ actions (e.g., installs) become significantly more valuable. If you’re trying to figure out what insurance company most of your friends use, whoever has the most Likes will be important to your search. And this doesn’t even begin to address the huge mobile discovery opportunity.
The second reason, utility, is a bit less clear but presents huge upside. We believe that Graph Search will help show users the inherent utility of the Open Graph. Before Graph Search, users did not have access to such robust data at their fingertips. Now that they do, people will start to realize how useful it is. This should translate into better discovery, increased Facebook use, and users more likely to participate in Open Graph than ever before because now they see the benefits. This means more app installs, Likes generated, and more. Not only will this teach Facebook more about its users and grow Open Graph penetration, it will also ultimately lead to more relevant ads.
Lastly, this sets Facebook up really well for mobile integration. By having Graph Search, users will be able sidestep the clunky mobile interface and type (or speak) their queries into Facebook’s increasingly sophisticated mobile app platform. And with outside search already handled by Bing (saving Facebook from having to devote resources to building their own search engine from scratch), they are ready to capitalize on the company’s biggest opportunity over the next decade. The full realization of this strategy is still a ways off, but the launch of a search product is a big first step for this play.
Ampush will closely watch and test Graph Search in order to see its impact on user behavior and the benefit it can bring to our clients. Make no mistake that this is an exciting product that will help users to quickly sift through all the content available on Facebook and discover new things within their social network. For advertisers, it will provide a way for brands to promote themselves on Facebook while leveraging the social proof that a Like provides to products.