Social Networks Set Sights on Search

Nov 02, 2009

November 2009 Issue

A series of recent moves has focused attention on search and social networks. Social networks have turned their attention to enhancing search features on their sites, but what they really may be searching for is the way to effectively monetize their services.

MySpace has long recognized the potential value of its search functionality, hinting that it might like to auction off its search function to the highest bidder. In 2006, reported that "Chief Operating Officer Peter Chernin, speaking during a question-and-answer session at the Deutsche Bank Media & Telecom Conference in New York, said that such a move would be one of the most lucrative ways to monetize MySpace."

This summer, things heated up. In August, Twitter quietly hired search industry veteran Doug Cook as its director of search-a move much commented on in the blogosphere. According to his LinkedIn profile, Cook has served as VP of engineering at Yahoo! and Inktomi. This addition to the team followed Twitter's VP of operations Santosh Jayaram's May discussion of the possibility that Twitter would not only index tweets but would also index the links included in tweets.

In an effort to improve its search features, networking giant Facebook acquired FriendFeed in August. FriendFeed originated many of the most popular aspects of Facebook's News Feed function, such as the "Like" function.

Why all the emphasis on improving search on these sites? "It's really pretty simple actually. ... They can't find stuff," says Susan Feldman, research VP of search and discovery technologies at IDC. Turns out it's a bit more complicated than that-but not by much. Like most things, it all boils down to money. According to Feldman, "You can't do ad matching unless you've got some sort of ability to match an ad to a query and to appropriate material. ... That requires that you have search technology." In other words, these sites need good search technology to make target advertising better and to build successful business models.

Any old search algorithm won't cut it because, as Feldman puts it, "Basic keyword search can only go so far." In an environment where all the content is user-generated, many problems can present themselves; not the least of which is the ever-changing language of cyberspace. "There are a lot of different ways to say the same thing," Feldman says, adding that keyword search will almost certainly have to give way to meaning-based search to really enable users to find what they are looking for when querying a site such as Twitter or Facebook.

Even with some of the difficulties facing social search, it has some advantages over a generic search engine query. Feldman says, "Looking in a smaller collection of information, you get better results." For instance, she says, conversations can be helpful. A user can find the answer he or she is looking for; but the user can also find some context. The original post may have good information, says Feldman, "but the remarks may offer a more-balanced view on that topic."

In essence, the real value to the searcher is the wisdom of the crowds. But Feldman also sees promising developments coming out of some of the old search staples. She says Yahoo! is worth keeping an eye on when it comes to the future of social search. Yahoo! allows users to opt in to a program that allows the service to troll your email, Flickr account, chat history, and other related data to better inform its search. As an example, Feldman says, Yahoo! could use her husband's information to realize that he is a coffee connoisseur and to know that when he searches "java," he means coffee and not the software. With all of the information available through user profiles on services such as Facebook, that kind of capability can only be a hop, skip, and a jump away.