Searching on Facebook: Why Google doesn't need to worry

<b>commentary</b> Before taking on Google, the social network will have to convince people to deviate from the only learned search behavior they know: looking for people. That's a tall order.

Jennifer Van Grove Former Senior Writer / News
Jennifer Van Grove covered the social beat for CNET. She loves Boo the dog, CrossFit, and eating vegan. Her jokes are often in poor taste, but her articles are not.
Jennifer Van Grove
4 min read
Mark Zuckerberg explains Graph Search at a launch event in January. CNET/James Martin

I'd really like to find my Facebook friends who like bacon and live in San Diego, said absolutely no one. The query is one of an infinite number of obscure searches that you can, but won't, use Facebook Graph Search to answer.

Graph Search is meant to be an intelligent search engine that helps people dive into the social-networking abyss and discover items long forgotten or buried behind the stream of information that flows by in News Feed. The "third pillar" product, which was unveiled in January and is now available to around 30 percent of U.S. English language users, is as core to Facebook as News Feed and Timeline, according to the company.

But there's a problem. Facebook faces the tall challenge of getting people to wrap their minds around using the new search engine for purposes beyond finding new friends to add to their network.

After nine years of not really searching on Facebook, why would we start now?

Facebook is tasked with the tall order of changing our behavior. All that talk about Facebook grabbing search dollars from Google is just talk until the company can alter the way we think about the social network and what it can do for us. Facebook was already processing 1 billion queries per day before Graph Search, but those searches, if you consider looking for friends a search, aren't really making Facebook money.

Ask Facebook why we will search, and the company will tell you that Graph Search is a user-defined presentation of the social network, where content is freed from company-ordained structures such as News Feed and Timeline.

In theory, Graph Search opens up a new dimension, one where every bit of social-networking information available to you becomes accessible on your terms. You can use natural language, instead of search terms, to happen upon people you may want to know more about, find places you might be interested in visiting, locate photos that would have taken forever to shuffle through, or glean what movies and music you might like.

In reality, understanding the language of Graph Search is far more complex than typing keywords into a search box, as we've all learned to do on Google. Because of Graph Search's odd parameters and its open-ended possibilities, navigating this second dimension is as unfamiliar as traveling to a foreign country without a guidebook.

Company watchers like Rebecca Lieb, Altimeter Group digital advertising and media analyst, and Jed Williams, senior media and technology analyst for BIA/Kelsey, wonder whether Facebook has waited too long in making the platform searchable.

Facebook has, in effect, trained people not to search, Lieb said.

People turn to Facebook to keep tabs on their friends -- and enemies. They come to boast about their accomplishments or rail against perceived injustices. They stay for an ego boost or to indulge in some ex-boyfriend stalking. They don't come to perform complex searches, because they haven't been allowed to, and they're especially unprepared for the conceptual search experience cooked up by Facebook, which requires an almost headache-inducing level of thinking to arrive at a query that returns information of any value.

Facebook is swimming upstream, Williams said, because people have come to expect one thing -- the feed -- from the social network over the years. "They're fighting the conditioning they've already done with their users," he said.

Loren Cheng, a product manager for Graph Search, won't come out and say that search is a foreign concept for the network's 1.2 billion members, but did hint that Facebook needs to hand-hold people in order for them to grasp what they can do with the new omni-purpose bar atop the page.

To address the what-should-I-ask-for conundrum, Facebook has consistently added to and tweaked the search suggestions it shows to people when they click on the bar. For instance, Facebook put in more photo-related suggestions because people who search for photos tend to get hooked on the product, Cheng said.

But the social network is playing an educated guessing game trying to determine what people want to do with Graph Search; it doesn't really know.

"We're looking at all the different audiences and trying to figure out what is [Graph Search] useful for," Cheng said. "I think that what we want to do is connect it with people's natural intent."

In an effort to make Graph Search more approachable to the average user, Facebook has altered the on-boarding tutorial and made 250 language-based changes to better ascertain what people want to find as they're typing their queries.

Like Facebook, Google has undertaken a similar massive reconditioning project, though exactly opposite in purpose, by appending a social network to its search engine. Two years later, Google+ claims 190 million active users, though its success is debatable and its high traction is due, in part, to a restriction that now requires people to sign up for Google+ when they create a Google account.

Without a similar tactic to force adoption, Facebook may have a hard time convincing people to deviate from the only learned search behavior they know: looking for people. Graph Search users query the most for other people, Cheng said.