Mar 17, 2009
The future of online search. Are we answering the wrong question?
Recently there has been some discussion about the future of search. Twitter have taken a position and made a substantial investment, but quite clearly Google, the giant of search, is unlikely to lie down and die just yet. But putting aside the technology arguments, perhaps it is worth considering what would make sense from the users perspective?
Quite obviously Google have done a sterling job of cataloging and indexing the HTML pages of text that make up the internet. Years of investment and some of the best talent in application development has allowed them to construct sophisticated models that derive ‘meaning’ from mere text. This is very clever stuff!
In their straight forward quest to provide the most relevant information to a users search expression, they have of course been slightly hampered by the rogues that would try to manipulate the results. But in reality this has been comparable to a flea on a large dog’s back. So given the progress that Google have made and the sophistication of their technology, what are Twitter presenting as their value and advancement?
Quite simply, Twitters case seems to hinge on the fact that Google have indexed “the past” (in other words published pages of varying ages), where as Twitter are making “the present” searchable. By making it possible to search the current conversations that are going on across it’s network, Twitter can give you an insight into what people are saying right now. No small task, and also likely to draw attention from the scoundrels that once tried to alter the purest search results on Google. After all is any harder to promote specific messages in Twitter conversations than it was to elevate a specific website regardless of its real value?
More of the same
However, anyway you cut it and as sophisticated as these applications become, you are still basically searching text strings in an automated fashion. To my mind that is fundamentally flawed and does not seem to represent any major advancement.
When a user searches for “web designers in Milton Keynes” they maybe looking for straight forward contact details (in which case text based search is fine), but more likely they are asking “which web designers in Milton Keynes should I be considering“. The latter question is much harder to answer. Or is it?
In fact what the user is searching for is a qualified referral. Something that is quite likely to be obtainable from across their extended network of contacts.
With the rapid growth of social media, an opportunity is starting to develop for a very new approach to search. One which leverages the power and “wisdom” of human beings across your social network. To make this effective and timely it is not quite as easy as simply tweeting a question or asking it through your Facebook status. A new search application is required that can make use of all the major “social utilities” and co-ordinate the results.
In addition, there is an opportunity to add real quality to the answer. Whilst using human researchers is not knew, the possibility of using trusted humans is a step forward.
Will this approach be safe from the vagabonds that try to manipulate search – of course not! Where this is monetary gain to be made out of laziness you are bound to find the “guru” vagabonds accumulating.
But at least we will be trying to answer the right question for the user.
Quite clearly this is not great news if you are Google. Unless you are already planning a completely new search application to run alongside your existing “encyclopeadic” database of past information. Interesting times ahead.





I wonder to what extent Google’s ‘Trends’ technology will help here:
http://www.google.com/trends/hottrends
After all, if people are chatting about it on Twitter, they’re almost certainly searching for it on Google…