Something that continues to puzzle me is the fact that Google's interface is so trivial. There are plenty of search companies out there aggresively innovating in the space of search interfaces. Vivisimo has cluster driven organization, A9 has multiple result types on the same page, Snap has interactive query refinement, result page previews and lots more. So why doesn't Google budge - or at least experiment via some other channel?
Then it occurred to me: Google relies on the momentum of the page rank brand (not the algorithm - it is buried beneath many layers of other relevancy tricks). The whole idea is that the user doesn't have to give much information to the system, Google will take care of everything for you. Google doesn't own any other search interface innovations, so if they started using new ideas here, they would expose themselves to risk from every other search engine out there. There would be a feeding frenzy on the next UI brand. By keeping their interface simple, they force the competition to compete with their brand.
There is another thing going on here: Google has made some strong suggestions that it believes AI (true, Turing-esque AI, not NPGs in Tomb Raider) requires huge data volumes to really exist. This view is something that Battelle reveals in his book 'The Search'. They very much like the idea that if users wait around long enough, Google will provide such a powerful system - enabled by all the information it has acquired - that there will only be need for a simple interface - the same simple channel that we use when, for example, IMing each other.
I have to say, innovation in the search interface is very hard. It is a lot easier to do nothing. And if it keeps the competition at bay - why innovate?
Comments