Searching for Lunch? Ask Hunch!

Hunch, the personal-decision-making engine created by Flickr co-founder Caterina Fake, launched yesterday to much press fanfare. The site has been in undercover beta since March, but in a blog post on the Hunch site, Ms. Fake introduced a new and improved version. “For those of you who are new, Hunch helps you make decisions, and gives you results it wouldn’t give other people, getting smarter over time as more people use it,” she wrote. “We would like it very much it if you would like it. We like it! So very much!”

Hunch users can create a profile and answer breezy questions from the “Teach Hunch About You” section. Queries include “Which fries would you prefer to munch?” and “Are you more likely to spoon or be spooned?” All of them were written by other Hunch users. Then users can ask the site questions (the search box includes a prompt: “Today I am making a decision about …”), and receive responses curated from their answers to Hunch’s questions and stats from other, similar users.

Some tech bloggers compare Hunch to Jason Calacanis’ Mahalo Answers. But that New York tech scene veteran’s “human powered search engine” is different from Hunch, which combines user opinions and machine algorithms to curate personal answers for its users. Mahalo Answers takes opinions from actual users and quotes their direct responses to specific questions.

“Hunch is wonderful (at first glance),” Mr. Calacanis wrote to The Observer in an email. “I just played with it for buying a car and it came up with a Tesla and Corvette in the top slots for me—I own both.

“In terms of a competitor to a knowledge exchange like Mahalo Answers I don’t think it’s comparable,” he added. “In a Q&A type service folks are interacting with each other, and in a system like this it’s rules based. I actually think they are complimentary.”

For example, Mr. Calanis explained, a Mahalo Answers user could give advice on buying a digital camera and say “check out Hunch’s digital camera test” to help out another user.

“These decision trees have been around for a while, but my guess us Hunch is going to do great because it will be, simply, better executed than the decisions trees on sites like PC Magazine and CNET in the past,” Mr. Calacanis added.

Ask Hunch, say, “What Should I Order At Shake Shack?” and the site will then ask 10 questions, including “How Much Are You Looking to Spend During Your Visit?” (“Under $5” or “More than $5”) and “What is the weather like today”? (“Winter-y,” “Cold and wet,” “Mild,” or “Hot and sunny?”). A results page, according to my answers, gave me three options (Single Shackburger, Single Hamburger, Hot Chocolate) and a “Wild Card,” “Pickles and relish make it special.” I can then like or dislike each result or write up pros and cons for each one. Another link to “Why did Hunch pick this?” will explain exactly why the engine made those decisions for me based on my answers.

Using the site feels a bit like playing a constant Choose Your Own Adventure game. Each answer to a question leads to a different result—building a “decision tree.”

Compare Hunch’s detailed Shake Shack results to a Google search for the same question.

Hunch also has other incentives: earn badges and credits, or “banjos” for contributions to the site. That’s right … banjos.

According to the site:

Hunch team member Caterina Fake has often talked about mainstream content vs. user-generated content being the difference between watching a television show vs. telling stories around a fire; or listening to Britney Spears vs. grabbing your banjo, going down to the parlor and putting the band together: Uncle Greg on the jug, Mom on guitar, Gertrude on vocals and Lucy on the mandolin. Dad supplies the bad jokes, the clapping, and the whooping. Given all that, we weren’t going to call them “Hunch Points” or something similarly uninspired. The Hunch currency had to be “banjos”. In retrospect, it’s obvious.

Combine those banjo incentives with a fun, easy-to-use site and Ms. Fake seems to be off to a good start. So far, many of Hunch’s users are from the early-adopter tech crowd, so topics and their curated results seem very, well, geeky. And some are simply … questionable, like “What gender am I?

A location-based Hunch service, like one specifically created for New Yorkers, would also be helpful for users.

But a combination of blog hype from Flickr devotees, not to mention a marketing boost from another new online “decision engine,” Microsoft’s big Bing, and Ms. Hunch’s site just might survive past the early adopter crowd.  Searching for Lunch? Ask Hunch!