Mahalo

From an uninterested, under-informed observer’s perspective, Dave Winer seems to be raking over old ground with self-publicist extraordinare Jason Calacanis over Jason’s human-powered search-engine Mahalo.

If you haven’t visited Mahalo, they basically take search terms like Guitar Hero III Cheats, New York City Hotels or MacBook Air and instead of creating software that finds lots of pages that mention those terms (like Google does), Mahalo gets a person to research the topic and write-up the best resources.

Jason’s talked about human-powered engines like this being better at deep-linking to the right kind of information than traditional search-engines do and that Mahalo’s here to compete with them. Dave’s not so sure and think that it’s Wikipedia that Mahalo’s got in it sights.

My personal view is that it’s somewhat a hybrid of of that. If you do searches in Google for specific things – people, places etc – you can see that Wikipedia is coming up in the top five. Indeed, 70% of Wikipedia’s traffic comes from search engines and 50% comes from Google. We also know that visitors who arrive at a site from Google are more likely to click adverts than other people.

Surely then, it doesn’t take a genius to work out that if you can get higher up Google’s rankings for a wide variety of terms (not just the generics that Wikipedia does well for) then you’ll get traffic from ‘searchers’ who are more likely to click on ads that take them to other (similar) places. Indeed, you should probably take the top 50,000 search terms from search engines and build out from there. That’s information that’s not always that easy to get, but hey, by billing yourself as a ‘search engine’ you might start creating your own data set that can help your progress grow.

So, is Mahalo a search engine? No. Is it trying to beat Wikipedia? No. Is it trying to generate as much traffic as possible to make money from contextual ads? Er, Yes.