Relevancy and the Internet's sea change
"(There is) increasing fragmentation of internet sites and users' desire to frequent specialized vertical sites, both in commerce and in content and community. This, in our view, is the last stage completing the Web 2.0 evolution which began with the decline of AOL as its value proposition of a walled garden rapidly faded away. We believe the new Web requires new models and those who do not adapt will eventually vanish. This online Darwinism, while in early stages, could entirely change the landscape over the next five years, if existing players remain slow to react."
At some later date I'll comment on my direct opinion of the above. That said, while I concur that the Web is undergoing a sea change which will require new models, I don't agree with Safra's perspective on users.
Specifically, I don't think users have a driving desire to frequent multiple specialized vertical sites. People still have significant limitations on their time, and users will still want to visit their favorite 5-6 sites per week (as they do now). The BIG difference is that in the past those 5-6 sites were focused on maybe a dozen branded destinations. In my humble opinion, that's the most significant shift occuring these days. The favorites are changing dramatically, as driving relevant content has become a focal point for new web sites/services.
Instead of a dozen destinations, we're talking literally hundreds of potential destinations that are making up the new 5-6 favored sites a user goes to an a weekly basis. Looking at traffic patterns of current top Web 2.0 sites, at least a couple hundred have 750K-1MM+ unique users per month. These sites are no longer concerned about crossing the chasm, but already have a deeply specialized cut of the mass market, driving specific relevancy to specific users. Take Dogster as a clear example with 300,000 registered members (and growing), the recently acquired Grouper with 8MM uniques per month and RSS Readers (like Bloglines) aggregating dozens of feeds into one site. Social networks based on age, demographic (college, adult), and individual idiosyncrasies (Goth, Music, etc) only extend this further. Personally I believe each 'relevant' niche will end up with a winner or two, but we're still far from judging who those winners will be. In the end, it's the huge propogation of relevant and free content/services from Web 2.0 sites which is driving this trend toward site disaggregation (all supported by ad revenue).
Every (and I mean every) major destination site needs to adjust and conform to this new model where that site's services are easily integrated to the NEW 5-6 sites now engaged by users, whether that's Bloglines, Dogster, MyYahoo, MySpace, Netvibes, or Youtube. At some point, consolidation will come (making it easier to partner with winners), but until then it's simply too risky to NOT wildly disaggregate and openly distribute online content/services.