free html hit counter A Brief Interview with Jason Calacanis of Mahalo | John Battelle's Search Blog

A Brief Interview with Jason Calacanis of Mahalo

By - June 12, 2007

Toro-1-1

Last week saw the launch of Mahalo, the new human powered search index from Jason Calacanis (that’s his dog, at left, one of his favorite online images). Jason has a flair for getting attention, and you all gave him some in the comments on my post noting the launch, all right. I emailed Jason for his response and we did the classic Searchblog back and forth. One thing I can say for Jason: He speaks his mind (see his comments on Squidoo, below). The interview:

What do you make of the response so far to Mahalo?



First off, I’m thrilled with the amount of attention and feedback the project has gotten. It’s a very contrarian idea, so I think some folks are rightfully exited and/or skeptical. Launching at the D conference certainly contributed to the excitement.

Of course, we’ve heard and developed a strategy to resolve the issues that people first think of when you say “human search” months ago. As you can imagine, when we spoke to our investors they brought up the same issues: how does it scale, how do you deal with human bias, how do you keep the search results up to date, why would people switch from google, how many search results do you need to reach scale, how is it defensible, etc.

I’m also excited that the SEO community has predicted it will fail because, of course, they have the most to lose when it succeeds! (Jason has been notorious for dinging the SEO community).

Why is this better than, say, ChaCha? And how is it different from the many approaches of the past that have not worked?



For the top 10,000 searchers we will be better than ChaCha because we have full-time, well trained, Guides spending 10 hours building each search result. That’s a lot of effort compared to the five or ten minute interaction with a live ChaCha Guide being paid, what, like $6 an hour (or $2 on a net-net basis from what I’ve been reading–ChaChat guides only get paid during ACTIVE searches from what I can see. So, if they get three five minute interactions per hour they make 1/4 of an hours pay for sitting in front of their computer for an hour–or like $1.50 or $2 I think).

That being said, I’m a fan of ChaCha for long tail searches. I like the concept of real-time searching and I think we’re going to add that feature to Mahalo.com at some point later this year. I’ve read many folks beating up ChaCha, but I think that ChaCha can make live search work in the way that Yahoo made Answers work after Google gave up on Answers.

Also, I think ChaCha would be a nice plugin to Mahalo, so perhaps there is a partnership opportunity for the two companies (although I haven’t reached out to them yet).

In terms of past approaches, I think Scott’s original vision for About.com/Mining Company *did* work, and the DMOZ absolutely worked–for a time. Both of those projects, however, suffered from neglect and a lack of resources. About.com has really switched from being a “guide to the web” to a destination. If you compare our Poker result with their result you will see they are trying to answer the questions and we’re trying to help you find the people who can answer the question. Both are valuable, but my feeling is that the world doesn’t need Mahalo to answer questions for them–especially not with the Wikipedia doing such a phenomenal job answering questions. The most we’re doing are the “fast facts” in the top right hand page of SeRPs. Those are like the top five facts to quickly confirm we got what you’re looking for, and maybe fill in some knowledge gaps users might have–not comprehensive in any way.

What the world needs now curation–not more experts.

Also, the world has changed in a number of ways since we last tried this type of project:

a) Software and bandwidth are essentially free–taking out 50% of the cost of running a web-based business. This trend is call Web 2.0… it’s gonna be big I think. ;-)

b) Google Adsense exists as a massive, scalable, and wildly efficient monitization engine. We’re not going to sell ads directly… we’re gonna leverage the services out there based on which ones perform best on a PER-SERP basis.

c) There are many more users online doing many more searches today

d) There is a lot more noise and “bad actors” polutting the web today (Blackhat SEOs, spammers, affiliate traps, splogs, etc), and as a result folks are frustrated with search.



Speaking of curation, what do you make of
Squidoo?

I just did a search for Paris Hotels on Mahalo and Squidoo, the result are obvious: Squidoo is a disaster and Mahalo is helpful. If you give folks the tools to make anything they will, and if you get some page rank the SEOs will come in and destroy your service. That’s what’s happened to Squidoo in the past year–it’s a dark, dirty, SEO back-alley now and I think folks are afraid to just say that. Do ten searches and look at the results–it’s garbage. There might be one good lens every once and while, but as a service it’s just horrible.

Seth had a good idea with Squidoo I think–empowering people. However, the execution was way too open and you can see what happens when you let anyone do anything without a mandate (think Wikipedia if it WASN’T an encyclopedia project and it didn’t have an active group of admins).

Squidoo is actually dangerous to use and it fails the number one test of Internet usability and trust: would you send your mother there.

That being said it has to be making a fortune for Squidoo and their ‘lens masters” because you can’t tells what’s an advertisement and what’s not. With Google’s new “let us know about deceptive advertising” program it’s only a matter of time before Squidoo winds up in the penalty box in my mind.

Seth should pull the plug on it.

… other than that I have not thoughts on Squidoo. :-)

You’ve said you can go four years without revenue. But your last business made a pretty good living on AdSense. Do you plan to do that again? Also, what about lead gen or similar deals? How do you plan to make money?



We are testing adsense already (see corvette page). My position, and the investors/board position, is that we should spend year one figuring out how to make *perfect* search results (a lofty goal), while figuring out how to scale our process.

As we all know search advertising is the most desired advertising on the planet because users are saying what they are interested in *right now*. It is the most effective advertising on the planet for this reason. If we build a great product it will scale, and if you have scale–as I learned at AOL–you can’t help but make money. Advertisers love scale–thats what federated media is hitting the ball out of the park–ya got scale you win!

Thanks, Jason!

Update: Jason just launched his Greenhouse effort to get the public involved in helping create search results…


Related Posts Plugin for WordPress, Blogger...

4 thoughts on “A Brief Interview with Jason Calacanis of Mahalo

  1. nmw says:

    Battelle hits homerun!

    Who was pitching?

    ;D nmw

  2. Andy Beard says:

    Hi John

    Thanks for the link through – I actually noticed it first when someone “repurposing” your content hit me with a pingback.

    I have actually given Jason some reasonable coverage on Mahalo, because it would be hard to criticise him for creating a website similar to things I have created in the past.

    I actually received some fairly harsh criticism for saying something positive about Mahalo, but then I am dealing with SEO for my own sites, and not those of clients.

    p.s. your intended link to Mahalo actually goes to ChaCha

  3. jfno says:

    Can someone explain to me the problem with the “Paris hotels” search on Squidoo? I just did it and the results were interesting, expect 2 about Paris Hilton. There was nothing preventing me to send my mother there.

  4. jackmayhofferr says:

    Isn’t Mahalo just a new Quiver? Sold to Inktomi back in 2002?