03 Jun 09 Why I’m almost Searching with Twitter

Over the past couple of years, a number of companies have launched human edited search engines (Mahalo, Wikia, Google SearchWiki to name but 3), and all of them have pretty much fallen flat on their shiny web 2.0 bottoms.  No real surprise there…

There is no denying that Human Edited Search is important.  By getting the results of the algorithm reviewed by real people, the quality should improve - and although Google don’t explicitly say that their results are massively influenced by the interaction of users, they collect plenty of information including relative CTR by position, bounce rate, and even engagement mapping / conversion tracking where sites have a Google Analytics implementation.

The problem with these various attempts at user edited search results is trust.  How can I trust an editor at Mahalo, or one of the mysterious and anonymous Wikia contributors to provide me with the results I want?  The simple asnwer is I can’t.  I have no idea what their agenda is, or whether their tastes are in any way in line with mine.  I simply don’t know them.

This brings me on to Twitter.  I follow people on twitter who I share interests with, whether those interests are work related (#seo), entertainment related (@jimmycarr), as well as family and friends.  These are people who I rely on to keep me up to date with things that are important to me in some way.  When I asked a question about Family Guy people responded and told me that it would be “this fall”.  They even provided me with links to pages about it, and one helpful chap gave me a link to a torrent file so I could download it - I haven’t.  This all happened in the space of around two minutes - not quite as quick as Google I’ll admit, but pretty good nonetheless.

This is all pretty cool, my twitter “friends” to give me information relatively quickly, and accurately, but the most important factor is that I can trust the information I get.  I wasn’t spammed with irrelevant websites as I would be in some search engines I could mention and I got what I was looking for.

Of course, the key word here is relatively quickly.  In a world where instant access to information is essential, relatively quick simply isn’t good enough.  Sure for now, I’m using Twitter to search because it works for me, but this won’t do for most people.

What I really need is some kind of better search.twitter.com that brings in the best of Wolfram Alpha - the ability to slice information from millions of users up into useful categorised chunks so that instead of just asking my limited number of followers for recommendations, I can instead outsource my questions to the 19 million or so Twitterers and take advantage of all the information that they have provided.

Tags: crowdsourcing, search, twitter