I had the pleasure of testing out Graph Search shortly after it first rolled out. First impressions count, and I must say that I’m quite impressed.
That doesn’t mean it is perfect – in fact, it is far from it. Facebook’s Graph Search is only at the initial testing stages now, and it will be using data from actual queries to guide future development.
It’s too early to tell who will win in the game of personalized search as Facebook gets into this (even Yahoo is getting into the game).
Facebook will be more natural to use for personalized results, as it has the advantage of being social by design. But its current Graph Search is rudimentary and not natural to the way people search (only certain pre-defined search terms are allowed, no “OR” and “NOT” Boolean operators, select from a pre-defined list of “suggestions”).
Google will have the advantage in “answer-driven” discovery search, with its rich experience in algorithms and organizing the world’s information. (But lacks the rich social information that lies behind Facebook’s walled garden. And please don’t talk to me about the potential of Google+ or I’ll start rolling my eyes at you.)
What’s cool: thousands of possible applications in your social life.
This is a bit of a read but follow me on this…
Let’s say I wanted to go to Turkey, and I am a little concerned about how a single lady should behave and what to look out for when I’m there. Using Graph Search, I find second-degree friends who have been to Turkey. In fact, I can view only people who are second-degree friends and who have moved there from Singapore (as opposed to people who’ve lived in Turkey all their life). I can refine this further by selecting only people who are currently living in Turkey (instead of just having visited the country), so I get the latest most relevant tips.
I have thus narrowed down the list to second-degree friends who have been to Turkey, and grew up in Singapore but are now living in Turkey who are female.
Through looking at their groups and associations, I pick out people who have mutual friends from my university and who listed that they’ve gone to my university (hence more likely to help me out).
Voila! Thanks to Facebook Graph Search, I now can reach out to them or our mutual friend for introductions, and I can have a better Turkey experience, thus saving money, time, and having better peace of mind. If this is possible for the average user, imagine the myriad of possible applications for marketers! (Imagine a Graph Search API.)
*Faces and names blurred to protect the innocent.
Many people have remarked that this spells the end of Google Search as all traffic will now be contained within Facebook with no need to move to Google for searching. This is taking it a bit too far and I don’t believe this will happen. What I do think will happen though, is a shift of response-driven advertising budgets to Facebook, if it chooses to monetize this feature.
To explain this: we know Facebook has already taken over top place from Google as the site where people spend the most amount of time. The only thing leading people outside of Facebook is to search for information. And this is where Google excels. In fact, I listed in a Search in a Social World presentation that a third of visitors to search engines came from a social networking site. If Facebook wants to stop this leak, it could develop Graph Search further for better results. A “search-like” advertising platform can be developed similar to Google’s AdWords – a platform that earns Google billions in ad revenue each year and sustains its other developments. Done right, this could be the answer to the long-term value of Facebook’s share prices.
Google isn’t resting on its laurels though. Perhaps some may not be aware, but the search juggernaut has been working on revamping search – one that promises giving more meaning, allowing for a more human way to discover information through understanding context. The goal is for computers to understand humans better. This is the Google Knowledge Graph project, led by Amit Singhal, who is responsible for Google’s ranking algorithms and rewrote the search engine in 2001.
Siri, despite all its flaws, suggests that people want computers to interact with them better, almost conversation-style, instead of looking at a data dump of information on a particular search for which the key term has to be refined over and over again before actually locating what they need. To give an analogy, people would rather have an assistant who recommends, rather than follow the signboards in a library or bookstore that lead them to the right book aisle.
One thing is for certain: the world will get even “smaller” as social networks and computer intelligence advances toward a more personalized experience for all.