Facebook formally announced their Google Search competitor, “Graph Search” today. The interestingly titled product (to go with Facebook’s Open Graph) is expected to roll out over the next few weeks across the world and will be in beta for a while. Mark Zuckerberg took pains to separate the product from Google Search but Google and everyone else knows better. This post covers todays announcement and my take on what it means for you. Like many of you, I have only seen video walkthroughs of the Graph Search and am stuck in the wait list waiting for my turn.
A few months ago, Google created a furore when it announced its integration of Google+ with Google Search. What this meant was that your searches would include parts of items shared on your Google+ feed wherever applicable. There were questions raised on everything from forced G+ integration to search being corrupted by not necessarily pertinent and high quality results. But folks in technology knew that this was a long time coming.
Social search has been the holy grail of search for a long time and Google had just taken its first steps towards integrating it closely with its traditional web search. It was always a matter of when not why personalized search was going to include our social networks.
Google also knew that Facebook was working on it. It was after all a logical progression of their vast troves of semantic data driven by our social connections and activities through their Open Graph system.
With all this happening, Facebook formally took off the wraps from its Graph Search feature. It is a significant step forward for Facebook as it tries to reposition itself not just as a social network but a social network that provides useful services. In one fell swoop, it entered the territory of Google Local, Google Search, and to a lesser degree Yelp!, Foursquare, LinkedIn and Angie’s List. It is definitely an ambitious step forward for the Menlo Park giant.
How will it work for me?
So what is Graph Search?. In layman’s terms, it allows you the Facebook user to search for things within the network. This could be for a good pizza place in Cupertino or previews of the latest Corvette StingRay. Results will be driven not by traditional web search crawls but by semantic searches within your friends and their actions on the network. So if your friend likes Amici’s in Cupertino and expressed it by liking the Amici’s Facebook page, that would show up in your results. You could then further refine your results to asking for pizza places in Cupertino that your friends living in Cupertino prefer and so on. You get the idea.
Ignoring the massive indexing and data mining challenge this means for Facebook (and hence the slow rollout and Beta moniker), the feature also requires certain things to happen for it to work meaningfully.
1. Your friend circle has to do meaningful actions to help in your search results. Going with the earlier example, if you have no friends who like pizza or specifically any pizza place in Cupertino, your search is pretty useless. Google Search or Yelp! can do wonders there.
2. You need a reasonably large friends circle. A small friend circle means a smaller trove of available public data to mine from and thus, not so effective results for your search.
3. People have to Like pages or leave indexable comments or Checkin at public places so that there is data that can be mined for recommendations. I could have 500 friends but if they don’t perform actions within the context of Facebook, it means nothing.
4. Businesses need to establish a Facebook presence and integrate with Open Graph system. This works very much in Facebook’s favor as it would drive more businesses to their services. If there are a 100 pizza places in Cupertino and only 10 of them have Facebook pages, the search is severely limited.
5. Facebook will have to go out of the way to convince people that they are not breaching their privacy. This is a huge issue. Facebook has not really been a beacon for privacy and the Graph Search is already making people leery. Google knows how bad it can get as they experienced during the launch of the Search plus your world service.
Potential of Graph Search
There is tremendous potential for Graph Search if Facebook can overcome the challenges mentioned above is a way that doesn’t cause a privacy firestorm. Semantic Search is the future of Search. People will increasingly lean towards search results biased by their social networks. The singular reason for the existence of Google+ is to gather data that is relevant to you- the videos you like in Youtube, the apps you have on your Android device, the songs you listen to, the places you go to searching via Google Maps and so on.
Facebook with a much larger user base to mine is taking what is really the next logical evolutionary step for the social network. Whether it is effective or not is dependent on the users and the small businesses that make the core of the network.
I will post my hands-on impressions of Facebook Graph Search when I get the feature enabled. Until then, here is a good walkthrough of the service. If you want to go behind the scenes on how Facebook built Graph Search with two ex-Googlers, read Steven Levy’s piece in Wired.