The other day Robert Scoble posted a couple videos on his thoughts about the future of search. Namely he argues that search has become so polluted with poor quality links that users are ready to look elsewhere. He asserts that the winners will be services that leverage social graphs to improve the search quality. His lists includes TechMeme, Facebook, and Mahalo as likely winners.
I'm not so ready to place my bet on these companies--outside of TechMeme, which has no other search service that's even close to providing a "searchable" view into the blogoshpere--but I do agree with Robert's suggestion that social graphs, or authority, can help improve search.
Yesterday he's posted links to a variety of responses--most strongly attack his assertion that search is in for a change.
I agree with Robert in that search needs improving. Searching blogs is one such example. The results are often poor. Blog's can contain a lot of worthwhile information, about personal experiences at work, play, in terms of health care, or eyewitness accounts of breaking news. I've pretty much settled on using blogsearch.google.com for most of my blog searching, but it is riddled with garbage--much of it spam blogs. I don't quite get it why it's so difficult to clean this up. TechMeme gives me a glance at what a pool of bloggers is talking about, but if you want to check on things outside of their conversational pool you have to go elsewhere. Currently there aren't many good "elsewhere" choices.
One point that Robert hints at is that if search at its most fundamental level was built with a social graph in mind, would it be much better. I don't mean strapped on. I mean at the DNA level of the search engine. I'd argue search would be much different. But this wouldn't mean it couldn't be gamed. I bet it would be. I'm also not sure if it would give me always what i'm looking for.
To me, the better place to create a search engine with a different DNA is to focus on what's new or what has changed. When I'm looking for authoritative archived responses that's one thing. However, much of the rest of the time I'm looking for what's breaking right now or what's been learned right now. It's a bit of a dichotomy, I know. Two extremes. That's why search can get so muddled I think when it's put together.
Robert is seeking more authority. I'm seeking better access to the latest information, organized by time--either most recent back or from original posting on a topic to most recent, or yes, maybe even filtered by authority.
Either way, I'd argue that at the core--at the DNA level--a successful search engine competitor should try something different. Robert puts forth leveraging social graphs. I'd build an infrastructure around deltas. Somewhere, sometime, I'm guessing someone is going to figure out a twist on search and when they do, we'll start changing, slowly our focus on search per se, to something else. With all the gaming of search engines today, I'm ready. Over time I think Robert's correct, others will too.
That being said Google could lead the way in this transition. There might be a trend in the direction of creating a cell phone optimized search that's even better at getting to what you want--in large part because the screen real estate and bandwidth are more limited than on a a desktop. Live Search has a mobile version that's thinking this way too. Who knows, maybe down the road we'll grab for our cell phones to do searches rather than a notebook or desktop because the large screen versions have too much extra "stuff" and the cell phones get me what I want, faster.


Incremental Blogger