Aardvark: Answering the Tough Questions

[ADMIN]

When Google was out pitching their business to VCs, the reaction of many was "search? isn't that problem already solved?" And, in many ways, it was. Yahoo was well established. AltaVista and HotBot had all the geek cred. And there were plenty of other search options out there. So why in the world would you fund another search engine? (answer: to get really really rich.)

Today, more than a decade after Google got started, one once again could reasonably make the assumption that search is a solved problem. Why would a VC invest in search when Google has virtually cornered the market? The short answer is that many VCs are deeply afraid of missing the next Google (and who can blame them -- Google was the best venture investment EVER). But that's a crappy reason to invest in search. (In fact, it is a crappy reason to invest in anything.) There are plenty of other reasons to look for yet another paradigm shift in search.

I believe that the best reason to continue to invest in search is that search engines are getting worse by the day. Why is that? For one, the amount of content on the Web continues to grow at a staggering rate. While there may once have been a mere handful of definitive sources for any given search, there are now thousands of relevant results for virtually any topic. That problem is exacerbated by the explosion of user generated content.

Far more problematic for search, however, are the economic incentives around the whole search eco-system. There is huge money to be made in search and all savvy online businesses are acutely aware of that fact. Because so much money is at stake, herculean efforts are put into gaming the system. Search Engine Optimization (SEO) has become an economic imperative for all businesses. And the object of SEO is not to get people the most relevant search results to their queries. The object of SEO is to drive the greatest amount of traffic possible to the optimized websites. In other words, the economic incentives of the search business assure that huge efforts are put into making search results less relevant, not more so.

Given those realities, it has been clear to me for some time that important new search technologies would have to emerge to help solve the "decreasing quality of search results" problem. Enter Aardvark. The Aardvark founders -- a group of entrepreneurs hailing largely from none other than the Google mother ship -- pitched me on the power of injecting human knowledge and relationships into the search process. By drawing upon the knowledge of your friends and their friends, the Aardvark founders surmised that you would be able to get more accurate, more relevant, better tailored answers to a huge range of subjective questions (e.g., "Where's the best place to eat sushi in Palo Alto?" "How can I best convert my VHS tapes to a digital format?" "I love The Decemberists -- any other bands out there that I should be listening to?" etc. etc.) Thus, the Aardvark team went about building the necessary technology to solve that problem, and I had the good fortune to fund them in that quest.

This week the Aardvark team is launching the fruits of that labor at South By Southwest (SXSW). They have built a "social search engine" that lives inside your IM and email. It allows you to ask questions of Aardvark, which then goes about determining who among your friends and friends of friends is most qualified to answer those questions. As the Aardvark team point out in their blog, Social Search is particularly well suited to answer subjective questions where "context" is important. Aardvark allows you to gather that context, both implicitly through the relationships you have with the answerers, and explicitly through the conversations between questioners and answerers. The resulting answers prove stunningly well-tailored to the person asking the question. And they avoid the pitfalls of the current search engines -- they are not subject to the vagaries of the proliferating user generated content, nor of the economic manipulation of search results.

I'm certain that there will be ongoing innovation in and around search. Getting the best possible answer to any question -- objective or subjective -- that can be arbitrarily posed, is a monumentally challenging problem. Aardvark goes a long way to addressing the shortcomings of search today and I am excited to see it roll out to a larger group of people.



More...