Posted by Ross Hill | 25 July 2008

Jason Calacanis

Mahalo

381 Comments

Tags: search engine

Jason Calacanis is well known for co-founding the Weblogs Inc network before selling it to AOL for $25 million in 2005. He is currently the “Entrepreneur in Action” at Sequoia Capital - who have funded Apple, YouTube, Google, Yahoo, Cisco, Electronic Arts, LinkedIn, PayPal, and many more.

Jason’s latest startup is Mahalo which apart from being Hawaiian for Thank You, is the world’s first human-powered search engine. All of the search result pages are hand-written by Guides who work in the Mahalo Greenhouse, while any searches they don’t have a page for are redirected to Google. It is these hand-crafted results that differentiate Mahalo from algorithmic search engines such as Google and Ask.com, as well as directory sites like DMOZ and Yahoo.

Mahalo seems to bring finding information back to the beginning of the internet but with a new twist. In the early days people used the Yahoo human-edited directory to find webpages and that was great for a while, but it was succeeded by the Google algorithm and then the Yahoo algorithm. Mahalo brings back hand-written result pages in the time of user generated content with sites like Flickr, Youtube and Wikipedia being hugely popular. How do you classify Mahalo and who do you see as your competitors?

I think we are a “search service,” which is somewhere between a search engine and a directory. We are not machines, but people using machines and social systems, and we’re not making jsut a directory–we are making search styled results.

So, we’re something new.

Not sure we are in competition with anyone but ourselves at this point.

You launched Mahalo at the Wall Street Journal’s D Conference, and then the Mahalo Greenhouse project which is where people can sign up to write search result pages at the NMK Forum in London. What sort of feedback are you getting so far?

Amazing. We’re getting tens of thousands of people a day already and we’re at 7,000+ pages created so far (each of those pages services 20 or so searches, so 150k searches are covered already).

The Greenhouse is exploding with activity. We have over 500 people registered and we’ve accepted over 500 search result pages (SERPs) in
the past three weeks.

On the hire application form you ask applicants for their blog address, their delicious account, their myspace account and more - all of the typical web2.0 places that are popular with the tech-heads. Mahalo seems to contrast this audience though and aim for the less technically-inclined audience who just want to get the information and move on. Do you view people with these online user-generated media accounts as the librarians of the web?

We think people who use Wikipedia, DMOZ, Delicious, Stumbleupon, social networks, and social news sites are very good at writing search results. That being said, we will accept anyone into the Greenhouse who wants to work hard and help people.

In June 2007 Mahalo had about 5,000 search result pages and since you are paying $10-$15 per result page for a person to write them it is obviously not practical to have a result for every long-tail keyword. “Australia” doesn’t even exist yet, but delegates the search to Google. How do you prioritize which pages are created first and do you plan to continue using other search engines to fill the gaps in Mahalo results?

We’re doing SERPs based on verticals and the raw lists of top searches. So, it’s not a perfect science. Frankly, it doesn’t really matter what we have now or don’t have right now–we’ll have all the important pages in a year or two.

On all the big news websites and in the tech blogs we constantly hear about what the American’s are doing but much less about what is happening around the world. You are based in Santa Monica, California but have recently traveled to Europe. As an entrepreneur who has been in a number of ventures what are international tech entrepreneurs doing differently?

The folks outside the US are too conservative and think small. Now, that means the chances of them flaming out are lower, but the chance of them building a Yahoo, MySpace, Facebook, Google, or eBAY level service are also lower. I think they should go for it more…