In search, nobody wants to go up against Google. Even Microsoft is trying to position Bing as a “decision engine” even though it is really just a search engine. But Hunch, a startup which launched publicly today, is an actual decision engine. The only thing it attempts to do is help you make a decision through a question-and-answer interface.
Hunch is part of a recent flowering of Q&A sites (such as Aardvark and Mahalo Answers) which address a part of search that is orthogonal to what Google does. In part, these startups are responding to the success of Yahoo Answers, but they also push beyond what Yahoo Answers does. Flickr co-founder Caterina Fake, who worked on Yahoo Answers before she left Yahoo and is now running product design at Hunch, says: “Yahoo Answers is not the answer.”
The problem with Yahoo Answers is that most of the answers aren’t very good. As long as one answer is good, that is the one which comes up highest in results. But it doesn’t really learn from all the answers to a particular question come up with a better answer. Fake explains: “The question has been asked and answered thousands of times. Collective knowledge systems don’t work unless they retain knowledge, you can add knowledge to them in a simple and straightforward way, and it gets smarter every time somebody uses it.”
I had a long discussion with Fake about what Hunch is trying to do on Friday. You can search for questions which have already been entered into the system, or add your own. Each question is actually a series of questions with multiple-choice answers which take you down different paths (or, “decision trees’). The questions and answers are contributed by Hunch’s community (so far the site has been seeded by its 40,000 beta users). “Anyone can add a new set of sub-questions to main question, and improve the overall results in the same way that many people can contribute to the same Wikipedia article to make it better or add tags to someone else’s photos on Flickr. It only takes aminimal effort to make each question marginally better.
For instance, the question “Should I get a convertible?” leads to a series of other sub-questions aimed at helping you make that decision: “Are you okay with the possibility that you’ll pay more for a convertible?”; “How’s the weather in your city/country?”; “Do you live somewhere known for car theft or crime?”; “Do you keep a lot of things loosely scattered around the interior of your car?”; “”Do you often wear fancy or personally valuable hats or scarves?” After you answer all the questions, you get an answer. This could be a yes or no answer or s specific recommendation, such as what blogs you should read.
At the end of this process, you tell Hunch or not you agree with the result. Hunch not only has a results algorithm, but also has a question selection algorithm, which it tunes to each person. Theoretically, the more questions you answer on Hunch, the more it knows about you and the better follow-up questions it can present to come up with the best final answer. It sounds complicated, but the user interface is simple and game-like. You are presented with a series of questions, and you click on the answers which apply to you. Then at the end, you get an answer to the original question you were exploring.
Fake estimates that about 40 percent of the topics on Hunch right now are monetizable with ads or affiliate links. If you try to use Hunch to figure out which camera you should buy, for example, and the answer turns out to be the Nikon D80, already you will see a sponsored affiliate link to Amazon. Other business models might emerge in the future.
Other sites will be able to tap into Hunch’s question-selection and answer-selection algorithms to create their own Q&A system. Using Hunch’s API, a developer could create a custom product recommendation app for retail sites. Bob’s Bait And Tackle shop could set up a series of questions and answers to guide shoppers to the perfect fly or fishing tackle. All of these questions and answers would then feed back into Hunch’s core system. The more people who use the system, the smarter it should get.
it takes forever to come to a decision with hunch! just try the ‘which convertible…’ part of the decision tree!
Is anyone else having issues loading Hunch’s site?
Hunch was pretty responsive a few hours ago.. looks like it’s having trouble now that more users are online.
Yah, seems the ‘crunch effect’ has hit it. Unless it’s coincidence!
I don’t think its the Crunch effect. I have been trying to use it earlier and its really slow and keeps going down.
It finally worked for me.
The homepage search box is nice. I typed in Macbook then hit “Should I upgrade my Macbook”. Pretty cool system.
Microsoft should buy it and integrate it with Bing. Maybe then Hunch would actually work the first time.
Notice how you didn’t say how good of an answer you were given.
That says everything.
I asked it what monitor I should buy. And it told me! But there was no reason why I should buy it. It just gave me a bunch of results. *Shrug* I guess it just gave me some leads to investigate. But unfortunately wasn’t very useful.
HTTP 500 error..after a few intelligent steps..
I will try later..
It worked now.. Interesting. However, I would still try to ask the question to simpler AAfter Search for other quick opinions…
she is in the unenviable position of having to spin her way through the process of justifying the existence of this odd web site. It is cynical to think that average human beings need to go to some web site to figure out if they want to purchase a convertible.
This is most certainly an idea in search of a market because there is NO organic demand for this product.
so true
You must be new to Web 2.0
“no organic demand for this product.”
there have been many inventions and business ideas that never had a real need or use…until it was used.
…and Yahoo!Answers is probably one of them, purely by it’s dynamics it became what it is, and if Yahoo hadn’t done it first then some one else would.
Why is she dissing it? probably and partially for publicity because she thinks she’s come up with what she thinks is the next and better service in the same mould, but the truth is, only time will tell. And users.
Caterina’s compliant about the Yahoo Q&A is like democracy: it’s a good system until you involve the people (read: don’t understand what’s going on and vote for the wrong candidates) and linking that back to Answers, I’d agree only on a level because a couple of the categories allow silly questions to be asked, and with silly questions come silly answers, and in the modern search everything web 2.0 some of the silly Q&A will be found.
*disclosure* I regularly browse Yahoo!Answers. I do take it fairly seriously, and try to answer the questions in an intelligent way – as do most users, but occasionally if the question is that in need of a funny answer I’ll give one. The category of P&S (regular users will know what this is) is probably most to blame.
.
^ it wouldn’t take msoft long to recreate this so is the acq even necessary?
True
I never tried to take any decision with the help of hunch but I like this service because it is really a decision making application and mind boosting.
I have a hunch that Hunch won’t be around long. It takes forever to get an answer to simple question like what web hosting provider I should use. At least with Yahoo! Answers, I can get feedback more quickly and often from people who know what they are talking about.
I also stumbled on another site with a different approach to YahooAnswers. http://www.qajack.com. The do qanda using video only, making the whole affair really human, social and fun. I guess we are seing a new breed of qanda site, which I think will bring more rich relevance to information.
I don’t know about that…Yahoo Answers often seems to have a bunch of a clowns on it that post crappy answers just for the sake of posting. And they’re often rude. I liked Hunch when I first played with it and reviewed it on usefultools.com, but then I found Crowdmind. I think it may be more useful when it comes to making decisions because you do actually get advice from real people:
http://alpha.cr...id=ewof0h86i8cb
Interesting facts & correlations
Hunch users who prefer McD’s fries need more than 10 min. to run a mile. But BK fry lovers like to play frisbee in the park. The 22% of Hunch users who are afraid of clowns
Not funny, I think that what gives “cool” web startups a bad name and hurt us all as an industry
The only conclusion I’ve come to is that I am very much not the target market for this sort of application. In fact, I’d say most geeks (or at least those of us who are quite good at research) probably aren’t. The real challenge is going to be getting this in the hands of the average person, who is really the kind of person who’d benefit from this.
Uhh…. Hello…. How is this any different/better than WikiAnswers or WikiCity’s Q&A platform?
As long as other people are doing things or at least telling me to do things, I will do them. Having no thoughts of my own is plus plus good
It’s an interesting concept, albeit a bit too cerebral for the mass market. Could be helpful for certain types of questions (medical, law, financial) – types of questions where you should really be speaking with a professional.
Cerebral? You’re kidding right.
hunch- a idea or belief based on ones feelings or suspicions rathan on clear evidence. example:this search engine was developed on a hunch.
My hunch is that no one will use it.
So, at her former employer aka Yahoo; she worked on a product called yahoo answers and learned about how it works and its deficiencies. Now, instead of solving those issues or sharing new ideas that would have resulted into new ideas; she quits and launches her own ‘decision engine’ that would compete against the very product she worked on. In addition, she publicly bashes the product she formerly worked on. Wow, she seems to have very good work ethics.
To be fair: You don’t know if she shared those ideas with Yahoo. Maybe she tried, maybe they didn’t listen, and that’s why she left. I would imagine there’s been more than a few businesses started that way.
CreateDebate.com does this very well, as well.
decision engines are overrated. stackoverflow and yahoo answers are good enough for me.
uggg….what a waste of 2 minutes.
i like yahoo answers better. Getting feedback from people that have some experience. sure there are some bad answers, but those are easy to spot and skip.
it’s pretty crap, no where near intelligent enough to make any meaningful decisions yet
Interesting site and article. Another niche website for question and answers for unidentified images is http://www.picanswers.com/
What tool should I use to help me decide which decision engine to use?
I tried it last night…. I won’t be coming back.
The number of “answers” they have is TINY and what they give you is of questionable value. You don’t get a straight answer to anything!
Either site isn’t the answer – with Yahoo I get medicore (and sometimes ’sponsored’) results, and Hunch takes forever to answer something simple.
She’s got a pretty face
Didn’t think much of hunch
Well, thats the idea.
She’s trying to prove that Flickr was not just pure luck when in fact, it was.
Hunch will be deadpooled in less than a year…. unless she pawns it off to some big corp.
There’s really no tech there… just a bunch of decision trees which are of dubious value.
Steve Marlin,
There is no need to bash Caterina. At least, she is out there doing something. I take my hat to entrepreneurs like her, like her idea or not, but people like her (in any domain) that keeps the world advancing. I criticized Caterina’s claim about Yahoo system (see my message at the very bottom of this thread), because I wanted to bring the readers attention here that Yahoo are already doing R&D is machine learning which is fact, but at least that I didn’t bash her.
Give respect to those who are out there developing something and risking their own money for freeloaders like you & me to use. You seem to bash every entrepreneurs who have been covered here at TC. You bashed Steven Wolfram and now Caterina Fake.
Have you developed a product that us readers here can check out if it is of superior to the products that you have frequently bashed here on this blog? If you do, then please let us readers know the link so we can check it out and see if it has no fault that deserves no bashing.
I give her respect for trying but Hunch is crap
btw, i heard that this is actually NOT her original idea. I hear that she got recruited by someone in NY to become a quasi-founder. I bet she ends up regretting this move.
Isn’t it too early to boast that Hunch is better than Yahoo Answers or even to compare?
Try “morning workout routine” in “Hunch” and “Yahoo Answers” and you’ll see what i’m talking about. Hunch’s suggested questions/search results were not even close to what I was looking for and it was rather distracting. Changed my keyword to “more workout routine”.
And I agree with John’s comment – June 15th, 2009 at 7:47 am PDT
In terms of business model if Fake is hoping the site will sustain itself off affiliate fees then we might compare Hunch to one of the many price/review aggregator sites out there e.g. pricegrabber.
With that in mind, I ask which am I more likely to turn to, for making a decision about a digital camera purchase? A site with a ton of reviews and historical price data, or a site that makes a seemingly arbitrary guess at what I might like?
I think the price aggregator sites will win out not only in terms of converting users (showing price / feature / review matrixes seems like it would convert more people than a computerized guess) and getting organic traffic in from people who want to make that sort of decision in the first place.
That said, I found Hunch fun to play with. I just think it’s an odd thing to aim for an affiliate revenue model with what they have.
I could of course be totally ass-backwards wrong. Lets see!
I think yahoo answers is simple enough. The “what blogs should I read question” lead me to TechCrunch and Mashable, these are blogs I know about already. There are 1000’s of other good blogs out there that am sure people in yahoo answers would have told me about.
There’s some potential for value in building recommendation systems like they’re trying to do at Hunch, but I think that it’s dumb that she has to degrade Yahoo! Answers to justify it.
Most questions on Y!A are a lot more complex than deciding what convertible to buy, and often don’t really have one specific answer. Often, people are just looking for the opinions of others. There are a variety of ways people use the service, and I’m not sure anyone who would put it down like that really understands it at all.
Fake? What a stupid name.
The site’s down, so none of this really matters anyways
Yahoo Answers is often criticized for bad quality of lots of the answers one can find there.
Is that true? To some extent it is, but given the scale of Answers (the number of questions and number of answers posted) it’s unavoidable.
I’m sure they already block lots of bad quality content and apply advanced filtering and ranking. Can they filter out all the junk – no they can’t.
You can think of it this way: in the past decade Google have redefined ‘relevance’ in web search. They have hired the biggest experts in that area and even if Yahoo and Bing are close, what Google shows is the state of the art. You can hardly do better. And still, they display lots of irrelevant results and quite often no relevant ones at all. Have they experimented with automatic question answering (as e.g. Wolfram’s alpha) – I’m100% sure they have. Shall we see it any time soon? I doubt – Google is too serious to launch an immature product.
Back to Answers: the situation is similar, you can complain but given the scale you won’t do better.
I bet at Answers they tried more intelligent scenarious (as the one hunch is based on) but for a serious company like Yahoo throwing every prototype at the users would not be serious, even if the ideas behind are great.
Fake did exactly this: build a prototype which is based on a nice idea but which won’t work that well in practice.
Answers still offers the best answer.
Here is the problem with Hunch. The following guy has a webdesign firm very dependent on Google for his livelihood, and voting Google up in Hunch. Talk about conflict of interest.
http://www.hunc...danlev/profile/
….what?
How did you make the assumption that I am “very dependant on Google” for my livelihood? I’m not sure I even know you, nor you know me.
Voting Google up in Hunch? Could you please clarify this? I don’t understand what you’re saying.
I prefer to take decision on my own, but its a rather nice decision engine. If you have the time to
answer 100 question just to get an answer!
Good Job, I like it.
I tried it out on a few questions and I can’t imagine ever coming back to use this thing. And I can sometimes be fairly indecisive. I found the whole Q&A to be tedious yet facile. I think Yahoo answers is a better solution by far. It reminds me of the ridiculous quizzes they put in magazines. But hey, maybe I’m not typical. I guess there is an audience for this. Maybe I should ask the question. Do people really care about decision engines?
Aside from the irony of answering multiple questions to answer one big one…
It seems like a cool idea and came back with some interesting, reasoned options for movies (the example I picked).
There are only about 14 people who REALLY worked on Answers at Yahoo! Everyone else puts it on their resumes (and the bigger their titles, the more likely they exaggerate their role).
Fake said…
The question has been asked and answered thousands of times. Collective knowledge systems don’t work unless they retain knowledge, you can add knowledge to them in a simple and straightforward way, and it gets smarter every time somebody uses it.
Fake described machine learning without her knowing that she did. I suspect that she has no background in it. Good on her for bringing her product to the market but one cannot hype up a product on the basis that is completely new (unbeknown to most internet readers), but actually I have no doubt that Yahoo researchers have already covered this area or currently developing it into their Q&A system. Why do I think this way? Well I have seen a few publications from Yahoo researchers that made it to various reputable computing peer review journals. What does this tell you? The R&D at Yahoo is alive and ongoing, so Fake’s claim is not right on the mark or at best it was an uninformed. A good example of this was an excellent paper from Yahoo Researcher, Sharad Goel who submitted a paper to the very popular machine learning journal NIPS (Neural Information Processing System) with :
Title:
—-
Predictive Indexing for Fast Search
Abstract:
——-
We tackle the computational problem of query-conditioned search. Given a machine-learned scoring rule and a query distribution, we build a predictive index by precomputing lists of potential results sorted based on an expected score of the result over future queries. The predictive index datastructure supports an anytime algorithm for approximate retrieval of the top elements. The general approach is applicable to webpage ranking, internet advertisement, and approximate nearest neighbor search. It is particularly effective in settings where standard techniques (e.g., inverted indices) are intractable. We experimentally find substantial improvement over existing methods for internet advertisement and approximate nearest neighbors.
There are many publications that have come out of Yahoo R&D, the one I cited above is just one of them.
It seems like a purchase decision making engine, as opposed to an engine designed to provide a more meaningful answer to common questions.
To me it just gathers user demographics and then attempts to sell you something, likely working as an affiliate for larger companies.
The problem with Yahoo Answers is perhaps the fact that it relies on now old-skool contextual advertising, while Hunch works on CPA advertising, which can be more profitable.
But who will want to answer ten questions every time they search for something, especially considering Twitter is now the norm for time spent on a certain topic.
I tried asking Hunch what sort of car I should buy. After answering 10 questions it advised a Honda Accord. Not bad, I bought a new Honda Accord last year.
Big Deal! The point is that you bought it without needing help from the Ouji board.
Nevertheless it would have saved me some time.
You had a preconceived notion of which car you’re driving. So when it asked you “do you like feature X”, you said yes because you use it right now in your car.
Now, had you taken this test BEFORE you decided which car to buy, you would have gotten a different result.
A better explanation is that I answered the questions the same as I would have several years ago. For example, one of the questions was, “Are you over 6 feet tall.” I was so preconceived on that one. Consider that I already own a Nissan Maxima which I bought in 2002. Very similar car. The word that comes to mind is “consistent” or “boring”, not preconceived.
Perhaps what I should be doing is asking Hunch if conversing with Steven Marlin is worth the effort.
Funny, even remotely suggesting that Mahalo Answers (a Yahoo Answers clone) does the same as hunch is hilarious. Where is the disclosure on this article?
they are trying to build an expert system out of user’s feedback. However, experts systems are known to be extremely limited to domain-specific knowledge, so, prepare your next ideas cause this is going to the deadpool
katerina was smart enough to guide flickr but not smart enough to realize when you get a payout, don’t worry about if people think you were just lucky or not, TAKE THE MONEY AND RUN.
instead she has listened to moron-critics who spend hours a day staring into the navel of web2 for no reason whatsoever, and she felt she needed to gain their respect as a for-real tech leader and legit entrepreneur. in the end she’ll probably have blown the wad she got from flickr and divorcing whatshisass, with nothing to show for it but a spot on the wayback machine
Sadly, rubbish. I asked it what running shoes I could buy. It asked some questions about brand preferences and then came up with a completely different brand as the answer to my prayers. How can you trust a site that is just going to put sponsored links at the top of the results whatever you input? #fail