Hunch_1262649667838Hunch.com, introduced to us in a early version last March, has demonstrated a growing user base and ever-better utility to go along with its fun-to-use design work.

For those not in the know, Hunch is a site designed to take some of the “cognitive labor” out of making decisions. Founded by Flickr co-founder Caterina Fake in December 2008, Hunch launched in June of this year. The site asks you questions about yourself and directs you towards products or decisions that others, who answered similarly, were happy with. Since it went fully public just six months ago, Hunch has grown explosively.

In an interview on January 4, Fake gives us some statistics just five days old. At the site’s launch, there were 30,000 unique visitors per month. In December 09, there were 1.2 million unique visits. Topics ‘played’ rose from 675,000 at launch to 7.7 million today, spanning over 6200 today. 52,000 user generated pro / con evaluations of the site’s results have also been made. In six months, Hunch has multiplied its unique visitor traffic 40x.

Staffing wise, Hunch welcomed Wikipedia’s Jimmy Wales to the fold on December 7. He will serve on the board of directors and as a design adviser. On his blog, Wales says he is impressed with Hunch and makes the prediction “Allow me to call it now: this is what we are going to come to call Web 3.0.” Hunch says they regard Wikipedia as the “gold standard to emulate… in community development.” Since then, Hunch hasn’t hired any new staff or raised additional money.

Bringing Wales on board dovetails nicely with Fake’s goal to make the site more social. She comments “we are planning some major user interface changes in the next few months. We are going to add a lot of social features for our users.” It is always a gamble to change a winning design. If the new layout successfully incorporates social content without alienating existing users, though, it can only further Hunch’s success.

Part of the decision making engine is in “tell hunch about you” questions. So far, 34 million of them have been answered. Hunch’s data set from its questions is huge. In a company blog post, Hunch claims to have found correlations between political ideology and culinary preferences. Amongst other things, the self reported data set indicates that liberals love Green Curry while conservatives eat more meat loaf.

It might not have taken $5 million in funding and a year of development to figure out that liberals like Indian food. Still, Hunch’s data set is a valuable marketing tool. When asked whether there were plans to sell the data, Fake said “No, we’re going to keep it for ourselves and maximize our users’ results.”

The technology behind Hunch is the brainchild of a group of MIT students. Using an algorithm to ask simple questions in greater specificity as you go isn’t exactly new. Using crowdsourced data from real live people to feed the algorithm, however, is. While details about Hunch’s founding are pretty hazy, it seems that these MIT folks contacted Fake with their idea for some development guidance. The group of them founded the company, which is based in New York City, went to an invite-only initial launch and has been fully public for six months now.

Hunch has competition from Yahoo answers, Wikianswers, Askville et al. In my experience, none of them are as fun to use or as helpful as Hunch. That, of course, is an opinion and I’m sure everyone else has them too. They might even be different from mine.

Fake, formerly of Flickr, is candid about Hunch.com in her blogs. Calling Hunch.com another decision making tool in the collection (which includes I Ching consultation and her sister), Fake acknowledges that some of it is still being made up. She notes, though, that while Flickr was more valuable as an acquisition than a standalone, Hunch has a strong business model. One worth sticking around for. At the end of our conversation, she says “We are all very excited about what is coming up. We have a lot in the bank, a good runway set up to take off from.” We look forward to tracking the flight.