alarm:clock

Search - Wednesday, May 30, 2007

I'm A Mahalo Editor. Kill Me.

We have spent time on two of the least scalable businesses we can imagine today. Everyscape is a neat mapping service that at launch had only mapped Union Sq. San Francisco. That's because it requires humans with panoramic cameras to took photos of every location on earth. That's babbling crazy talk.

But Mahalo, the new search engine launched with much fanfare at the D event today is One Flew Over The Cuckoos Nest Meets The Shining crazy. Calacanis has hired 40 editors to edit and format the top search terms on the Net and work down from there until they are done. Wikipedia is edited by thousands of people, can you imagine as an editor being tasked with editing 1/40 of the Net? (Update: The deal has a funny post from a writer who had to edit at a similar startup that preceeded Mahalo called Magellan.)

Mahalo's stated reason for being is that SEO has tainted the purity of results on Google searches.

In the mind of Calacanis, we can guess, this is not that different from Weblogs Inc., which hired editors to keep readers up to date on topics like gadgets, travel and health that are top points of interest. And Calacanis make that point that the top 10K search terms account for a quarter of all searches.

The trouble is that Calacanis is marketing this as a search engine and as such it returns disappointing results almost all of the time. If Mahalo hasn't hand-crafted a page for your search, it tells you so and presents a few custom pages that are close, and below that you get Google results. We buy Calacanis' data point about a cluster of top searches, however, another key data point is that almost every does long tail searches.

Also, if your expectation is that your getting hand-crafted, you might expect that you are probably going to miss some of the more recent data, so after doing a Mahalo search, you'd likely want to search Google and or Technorati as well. So as a user, my expectation is to expect failure and be prepared to scroll down. That doesn't hold together.

To research a vacation, we searched on "Vietnam vacation" that didn't work so we tried "Vietnam." That didn't work either so we clicked on Richard Nixon. The first thing we saw there was a link to Wikipedia, which seems like a better source for information on Richard Nixon than Mahalo.

As a business, we can't imagine how Mahalo succeeds. Sure it will take revenue but it has to be a staff of 40+ editors, plus techs. Calacanis tells TechCrunch that he has enough money in the bank for Mahalo that he can stay afloat for 4 years without revenue. To his credit, Calacanis also acknowledges the weakness of Mahalo saying that it “It takes a few years to build a really compelling search engine.”

Funding for Mahalo came in two rounds. The first was led by Sequoia Capital. The second was led by Elon Musk and News Corp. Other investors are CBS, Burda Media, Mark Cuban, Fred Wilson, Ted Leonsis and Jonathan Miller. PaidContent reports that they have heard the funding totals $16M on a $100M pre-valuation.

Thus far Mahalo has generally received strong marks from reviewers. Michael Arrington writes: "Mahalo is a search engine, and will join Powerset as the more interesting new engines to launch in 2007." Rafe Needleman at WebWare writes "It looks like a very useful service." But if this wasn't Jason Calacanis' company and if Sequoia and News Corp. were not in, a true blind taste test, we have to believe they would have panned Mahalo and never gone back.

View - site
Read - Jason Calacanis Launches Mahalo Today: Human Powered Search