With Buzz, Google as a search network

Google thinks that now is the time to incorporate the social-network into the search engine | IslamOnline.net

Perhaps surprising to some, Google too makes mistakes. This fact was on display with the February launch of Buzz, Google’s social networking service apparently aimed at competing with Facebook and Twitter.

The problem was not Buzz itself, but the hasty manner with which Google pushed it onto its Gmail users. Like Twitter and Facebook, Buzz enables users to post “status” updates, links, photos, and the whole repertoire of Web-borne distractions.

However, eager to make its presence immediately felt in the new realm, Google decided not to start from scratch. Instead, it tied the new social-networking service to Gmail, its mail service with about 175 million users. And before users knew it, they were, without permission, signed up on Buzz, and assigned lists of “followers” based on the frequency of email exchanges with people in their address books.


Since “followers” could see each other, Buzz revealed information about users’ contacts (for instance, their doctors) that they most certainly wanted unrevealed. Making things worse, at least initially users could not turn Buzz off nor sign out of it. The result was a buzz of complaints (with at least one court case) followed by an apology from Google.

Of Networks and Engines

The Buzz controversy raises two questions. One is why would Google, a firm known for letting its services brew in the “beta” (or the-service-is-being-refined) phase for years, rush a service potentially as important as the Internet giant’s answer to Facebook and Twitter? After all, Gmail and Google  News, to note just a couple of examples, stayed in beta for about five years — five years, that is, after they were released to the public.

The other question is why does Google, whose stated mission is to “organize the world’s information” enter the social-networking fray? The link, one could say, between a social-networking service and
organizing the world’s information (or the search engine, Google’s core strength) may not be immediately clear.

A little-noticed announcement from Google in the immediate wake of the Buzz launch probably offers the answer to both questions. On February 12 — that is, three days after rolling out Buzz — Google let it be known that it had acquired Aardvark, the social search engine.

In Aardvark, the code of the search engine is designed to pick the right individual (in a user’s network of acquaintances) who is more likely than others to answer a question from that user. That is, while existing search engines help users by indexing and ranking documents, Aardvark indexes and ranks individuals based on their expressed (or discernible) interests as well as their social-networking links.

The Aardvark acquisition thus throws light on both Google’s Buzz haste and its decision to offer a social-networking service in the first place. Google, through Buzz, seems to have wanted to have in place a platform to implement the novel “social search engine” concept before somebody else does.

As such, Buzz (which will somehow incorporate Aardvark sooner or later) is not directly aimed at competing with Facebook or Twitter. The Buzz-Aardvark duo appears to be essentially a search, rather than a social-networking, service.

It will still be competing with Facebook, of course, but fundamentally not on social-networking grounds (as in, say, which one is better at embedding videos into “status” updates). Rather, the mode of competition more likely to ensue will revolve around this question: which social-network is more capable of making the social connections in a network useful in terms of the information needs of each member in that network?

Put differently, which social-network will be first to harness the so-called “social graph” (the map of connections between users) towards the goal of finding, organizing, and ranking information?

From the Library to the Village

That — to harness the social graph — is apparently what Aardvark aims to accomplish. Granted, a service in which a user asks a question and others provide answers is nothing new. Yahoo and Amazon, among others, have been offering just that for several years. Aardvark, however, has come up with a code-mediated mechanism to rank answers (using the strength of social connections as well as other criteria) in a manner that resembles how Google ranks Web pages based on the number (and weight) of the links they receive. A weight of a link is determined by the popularity of the page from which it originated.

As Google embraces the social search engine concept, it actually hints at two intriguing trends in how people deal with information. One is that code is insufficient. Google may construct and maintain admirable databases, or libraries, of documents, but that still does not guarantee reasonably effortless answers to users’ “queries.” In fact users put in “keywords” rather than the more natural “queries,” which makes getting straight “answers” from existing search engines variably demanding.

Because of that, Damon Horowitz, founder of Aardvark, says in a co-authored paper that his “search engine” employs the “village” rather than the “library” paradigm. In other words, On Aardvark information is sought through direct interaction between the “asker” and the “answerer,” rather than by querying a database as in a library.

The (Social) Network is the Computer

The other trend is that a new breed of services stands on the assumption of real-time, all-the-time connectedness. Aardvark is an example. It assumes that the best “candidate answerer” will be online to respond to questions. I asked Horowitz if that is a valid assumption. “ Aardvark works especially well when delivering answers from people who are online over [instant messengers],” he said.

“But many people also answer questions from the vark.com website, from their iPhones, and over email -- places where we do not traditionally think of them as having "online" status. People are more available than ever today to answer questions through digital communications channels,” Horowitz told IslamOnline.net by email.

In this sense, every Web user is a “node” in the network -- similar to a PC -- and potentially has information that others may deem important.

And that is part of why, the poor privacy controls aside, linking Buzz with Gmail is an intelligent move. It piggybacks on an existing information connection — between those who exchange bits of information via email.

In the meantime, the privacy woes that accompanied the Buzz launch probably served as a useful reminder to Google that as the lines between the network (computers) and the social-network (individuals) get further blurred, it will need to take privacy even more seriously.

---