Why Facebook search won't be a Google killer - yet
"You will never make that work, but if you can it'll be awesome" ... Facebook CEO Mark Zuckerberg. Photo: Getty Images/AFP
Facebook's Graph Search holds the promise of providing instant recommendations for restaurants, cafes, books, movies, even dentists, all based on what your friends like, but it will be a while before it becomes a Google, Yelp or LinkedIn killer.
Facebook founder Mark Zuckerberg calls Graph Search the third pillar of Facebook, after Timeline and the News Feed. Initially focusing on people, photos, places and interests, unlike Google, Graph Search can answer questions such as "Restaurants liked by my friends in Sydney" or "TV shows my friends like".
But as early reviewers such as Engadget and Gizmodo have found, the search results are fairly lacking unless friends spend the time extensively updating their profiles with all of their favourite restaurants, hobbies and likes.
Facebook Graph Search
Photos from your search are displayed in the order in which Facebook thinks they would be most important to you. Importance is determined in part by the number of likes and comments a photo has.
"Unless your friends are pouring opinion and activity into Facebook, searching can feel hollow," Gizmodo wrote. "And until Facebook starts incentivising that kind of annotation — or making it uncomfortable not to — search will continue to feel kind of vacant at times."
Graph Search is the brain child of a team led by Google engineer Lars Rasmussen, who Facebook poached from Google Australia in 2010. Previously, Rasmussen built the first iteration of what became Google Maps and he was also responsible for the now defunct Google Wave.
In a blog post announcing Graph Search, Rasmussen uses "Dentists my friends like" as one of the three examples of search queries he gives in his first paragraph. This assumes, of course, that people are "liking" their dentists on Facebook.
Rasmussen sits on the board of Australian social search start-up Posse, which helps Facebook users "discover your friends' favourite places".
Posse founder Rebekah Campbell perhaps best outlines why Facebook search will be an uphill battle.
"[The] problem they have is their data isn't very good, it's big, but it's not quality," she told Fairfax Media.
She said while Facebook Graph Search was useful for finding things such as friends who have completed an MBA at a certain university, it won't help people find great coffee near them.
"Hardly anyone checks in [to places] on Facebook and people who do aren't particular about whether a place is good or not — they generally check in to show what people they're with — not talk about the great place they're at," she said.
"So if you run a search on Facebook for coffee what you'll see is where people have checked in that you know, not where you're friends actually love because people don't tell Facebook that information."
The Facebook search tool also isn't integrated with services such as Yelp and Spotify yet, which could go some way towards helping it offer better recommendations.
Slate's Farhad Manjoo, who bluntly states Facebook's "search results aren't that good", illustrates the data issue in his blog. He searched for "Restaurants liked by people who live in Palo Alto, California", and the result he got was Facebook's corporate cafeteria, the Facebook Culinary Team.
"Now, I've had a couple good sandwiches at Facebook's cafe, but I don't remember ever being knocked off my feet," Manjoo wrote.
Search Engine Land had similar issues when searching for restaurant recommendations.
Analysts have said they do not believe Graph Search will be an immediate threat to Google. Some even argue it will be good for Google as it now has more support for its claim to anti-trust regulators that it does not have a search monopoly.
After years of collecting all manner of data about our lives it is ironic that Facebook needs more intel to get social search to work. The social network itself admits that it has years of work ahead to realise its potential in this area.
As Rasmussen says in his account of his meeting with Zuckerberg in 2011 when he first showed him the idea, Zuckerberg told him: "You will never make that work, but if you can it'll be awesome".