The news broke today that Google will be buying Aardvark, a human (and algorithm) powered social search engine that I have written about quite a bit (early last year, most recently, all). I’ve also featured the service’s founders at both Web2 and the CM Summit.)
I’ve confirmed the news in an email with CEO Max Ventilla.
I can’t say I’m surprised by this news. Aardvark’s founders and advisers have strong ties with Google (Ventilla worked there, and a key adviser was at Kaltix, which was purchased by Google).
To me the critical question around this move is this: Will the Aardvark acquisition be a Dodgeball, or will it be a Applied Semantics? With Dodgeball, Google bought a promising startup in a strategically important space, but instead of integrating the technology and committing, it let it languish (the founders left and started Foursquare). Google later determined it must play in the space, and rolled out any number of features inside its mobile, map, and even Gmail products that mimic Dodgeball’s early features.
With Applied Semantics, Google again bought a promising startup in a strategically important space, but this time it successfully integrated the company’s technology and team, driving a crucial new business – AdSense – to become a critical and game changing business for the company.
So which is Aardvark? I’m not sure anyone at either company is sure, but Google is spending a reported $50 million to make sure no one else can find out. I do know Max well enough to say that his goal would be to see Aardvark integrated into the main search interface, such that when you ask Google a question, it would give you the option of “asking a human” through the ‘vark service.
Now that would be pretty cool.
Not to mention, ‘vark uses Facebook Connect as its core social graph for question answering. I certainly hope that will stand as the company integrates.
The parties can’t speak on the record about this yet, but another hope I have for this acquisition is that some of that ‘vark DNA about humans being critical to search – not as data points, but as part of the solution, connecting one to the other – will somehow infect the Google genome. We’ll see….