Mahalo is the world’s first human-powered search engine… apparently

Are people better than computers at knowing what you want?

Or rather, more importantly, does it scale?

How many people would it take to compete with Google’s data centers full of computers? Which could create the result to a search engine query quicker? How about more accurately? While I would wager a lot of money on the computers winning, there are some who are not so sure that computers are the best at this sort of thing.

A new search engine called mahalo.com has been created that has manually created results. It’s basic plan, as far as I can tell, is to provide better results for the top 10000 search queries than Google or Yahoo! To achieve this, people are going to manually create the SERPs. Here is a list of possible issues:

Freshness of results.
Bias in results.
Results that haven’t been created.

As you can see, these are all issues that a certain open directory called DMoz has/had. So is it going to be another DMoz? At the moment, I think the answer is no. However, I also think that calling this a search engine is wrong as well, but more on that later.

As far as I can tell, the editors are all paid, which means that hopefully, they won’t have an agenda, and will be unbiased. But as we all know, people as a whole are naturally biased by nature, and there is very little that can be done to stop this. People also don’t scale as well as computers. While 2 people might do twice the work of one person, 500 people won’t be doing twice the work of 250. You get increased overheads (ie. you need HR people and all the other magic stuff that larger companies need) and you cant get everyone to work at the same skill level. People also don’t work 24 hours a day. Or 7 days a week. Or holidays.

So what we have is lots of little algorithms (people) that are all different, going around creating SERPs as fast as they can. On the other hand, we have Google, with bazillions of computers in data centers, visiting billions and billions of pages everyday and creating SERPS on the fly using just one set of algorithms that judge all pages in the same way. Which would you trust to do a better job?

I said earlier that I didn’t think that Mahalo.com is a search engine, and I suppose I should explain myself. In my opinion it is a list of sites about the most popular things on the internet. A search engine is not a most popular things on the internet site, it is a site about everything. If there is a website related to the search query somewhere on the internet, it must find and return that page in the SERPs. It cannot say “Oops! We haven’t hand-written a result page for “snowboarding monkey” yet.” and then return results from Google.

I think that this is an interesting project, but not one that I would give much thought to in terms of SEO or anything - if you are an authority site, and you are liked by the editors, and you don’t have too many ads, and you meet all the required criteria then you will get in, if you aren’t you wont. At the moment it doesn’t matter. And it probably never will.

James
SEO Programmer

Listen to this podcast

No Comments

No comments yet.

RSS feed for comments on this post
TrackBack URI

Leave a comment