• by Laurie Sullivan  @lauriesullivan, April 1, 2014

Facebook claims no connection between its work on Graph Search and the company's new crop of search engine marketing experts, but the recent group of new hires like search expert Frank Lee from Datapop suggests otherwise. Its focus on better understanding and perfecting semantic Web search will give it a stronger position in mobile.

Quite a few Googlers and search experts from other engines have transitioned to a variety of positions at Facebook, according to sources. More than 100 Facebookers work on the Graph Search team, which also supports the site's traditional search product.

Reports put Graph Search in the hands of some marketers on mobile, but Kate O'Neill, Facebook product marketer on the Graph Search team, said the mobile version is really only available to a handful of users testing the platform.

The news of the small tests on mobile Graph Search generated lots of hype after Facebook reported record quarterly results in the December quarter linked to a 76% increase in advertising revenue, compared with the prior year's quarter.

The tests are "small," but in aggregate represent a "pretty significant navigational change in terms of how people get to information," O'Neill said. "The type of search we have on the Web through Graph Search doesn't translate well to a small screen, so we're focused on building out the experience on the Web first."

Facebook and Google represented the majority of worldwide mobile ad market growth last year. Combined, the two companies generated $6.92 billion in net mobile ad revenue in 2013, per eMarketer.

Changes in Facebook's infrastructure continue to threaten the organic reach for brands. Research from Ogilvy & Mather suggests companies have dropped from 16% of followers in 2012 to 6% of followers in February 2014. For large pages with more than 500,000 Likes, organic reach hit 2% in February. Some believe the channel will turn more toward paid.

Today, paid advertising options do not exist in Graph Search, although Facebook provides a guide to optimizing organic content. O'Neill said brands need to keep their company information on Facebook pages accurate and up to date, which will support Graph Search as a discovery channel.

Facebook engineers have been working to build out a search system called Unicorn for more than three years. It was designed to quickly scale all basic structured information on the social graph and to perform complex, personalized operations on the results. Unicorn resembles traditional search indexing systems and serves content from memory, but it was built to support social graph retrieval and social ranking. It is the primary back end for Facebook Graph Search designed to serve billions of queries daily within milliseconds.

O'Neill calls Graph Search "an extreme technical engineering challenge" because it understands short, simple phrases and respects individual privacy settings. The engineers have built the platform to recognize social signals, which pulls in unique data based on the person searching for information. The content returned is based on the searcher's connected network of people and the information they share.

Lee and others at engines have been making the leap to social to take on the responsibility of managing social budgets, along with their search budgets. Havas Digital in 2013 created a social division in Europe called Socialyse, which the company brought to the U.S. this year. Search engine marketing expert Jim Grates heads up the Havas Digital Socialyse division for North America.