2/16/2003

Google buys Pyra: advantage, UNMEDIA!

It's awesome that Google is buying Pyra Labs (the folks who invented Blogger). Arise, Bloggle! There are already naysayers, who argue that Google's focus on search engines doesn't fit with the blogging paradigm. But I think this misses the point about what Google is, fundamentally. Google says it's all about "search". But "search" is really just marketing-speak for data mining. With blogger, Google gains a human-driven, massively distributed, real-time updated store of content that is enormously useful. This is a deep, rich vein for Google's datamining technology.

In fact, I had actually written to Google last October suggesting that they focus on weblogs! The general idea was :

Your news service is brilliant, but it made me wish for another similar feature aimed at weblogs (and tied into the news service also).

What would be wonderful is if you could do almost exactly the same thing you do for news feeds, but using RSS feeds from weblogs. Currently there are hundreds of thousands of weblogs run on Blogger Pro, on Radio Userland, and on MovableType, all of which support XML syndication. It should be easy for your team to build up a comprehensive list of weblogs by looking at the recently updated list at weblogs.com and blogs.salon.com, as well as
blog.gs. Blogs with XML feeds could be flagged and used as information feeds into the same algorithms you use to construct Google News.

With that data, I propose you do two things:

1. create a weblog-centric normal Google search
2. link weblogs to the news.google service so that if a number of weblogs link to a given story, users of Google News can immediately find those bloggers commentary.

Weblogs are posing a significant challenge to the normal media and often contain intense, detailed discussion and analysis of current events as well as general debates on politics, religion, the economy, and foreign policy. All of these diverse threads of opijion and analysis could thus be indexed and accessible via the Google News interface.


There's more to the letter, as well as Google's (boilerplate) reply, in my original post.. As Glenn would say: Advantage, Blogsphere!

UPDATE: Cory Doctorow has also devoted a lot of thought to this, and has more ideas on how Google could leverage the vast content resources it just acquired. He points out that Google's vast volume will allow it to rapidly innovate and the Blogger data will become a custom test-bed for Google's idea laboratories:

But it's that usage-volume at Google that makes this deal so exciting. Like Amazon, Google has so much traffic that it can afford to roll out small-scale trials -- Remember the thumbnails of search-results? The limited trial of Folding@Home in the GoogleBar? -- and get instant results about how well a new feature performs. Google's core expertise is making sense of data gathered from the Internet, so it's eminently capable of making sense of the results of these trials.

What this means is that once Google actually does integrate Blogger proper into its service, we can expect very rapid and very solid innovation. Gbloogle will be able to sneak features in for a day or two, extract the data, and make some sense of the data, decide whether its worth keeping the feature, and engineer something Google-grade to put on the back-end.

No comments: