Jason Calacanis has pulled back the curtain on his new start-up, Mahalo, an alt search engine that pairs search queries with results from humans (paid and unpaid) as well as from bots (Google.)
At first pass, I mistook the new service as an effort aimed at returning hyper relevant results. After playing further with the service, however, Mahalo appears to be far more intent on displaying what’s popular than on what’s relevant.
Hand-written results pages and categories make me think of the Yahoo! of last decade. I don’t see how that’s sustainable or valuable. And while social participation (especially in the form of community ranking) has a very important role to play in the maturation of content discovery, the user-submitted links and popular searches featured on Mahalo are pushing down the wrong path.
Rather than enhancing what already works, they’re trying to bolt on more. More results, more related content, more user-submitted links. As far as I’m concerned, more is the opposite of better.
I’ll stick with the algorithm for now.
Where I could see services like Mahalo or ChaCha adding value is in the form of a toolbar widget that provides expert, human suggestions for search queries upon request (similar to how sphere matches posts of similar topic.) As destinations, though, I don’t see long term value in the sites themselves.
Anyone on the other side of the fence? Am I missing something?
One thing’s for certain, Mahala.com is getting a lot of traffic right now.
8 Responses to “Who needs algorithms?”
Obviously several VC guys are banking some money on these human-generated results or whatever. I have yet to use one that I thought was all that great (including mahalo). So, I’m with you, I’m also stuck using Google for the time being. //g
I’m tempted to say that something like this would never scale, but I said the same thing about Wikipedia back in the day.
If Wikipedia paid a team of experts to populate the entries, they’d have problems scaling. Not only does Mahalo need to write and categorize, they need to prune bad content / links for as long as they exist. That’s a decent amount of work.
Who knows? Maybe the internets will get smaller and become more manageable.
[...] like only yesterday I was saying that “social participation (especially in the form of community ranking) has a [...]
To split a hair, “what’s popular” is the outcome, not the intent. The intent is spam-free safe search for those who don’t know how. It’s hard to tag this kind of target audience without making Mahalo sound like internet for idiots, but lots of people submit generic Google searches and click random results. Mahalo wants to be the site recommended to aunties and grandmas frustrated by Google. It’s remedial search.
It’s tough to say what their intent is. But even the fact that “Today’s Top 20 Searches” are featured more prominently on the results page than anything else (including the search results themselves) demonstrates that Calacanis has brought his background into this project.
You make an interesting point about Mahalo’s (secret) target audience being non-savvy grandmotherly types. I didn’t know those people got frustrated with Google, though – I thought they spent their time at Yahoo!
y! or msn :) //g
[...] RSS can help you make sure your audience sticks with you whilst you a) grow the content base, b) make your objective for this site a little clearer and 3) turn drive-by’s into trusting subscribers. Give me a feed for the [...]