Thoughts on a Day With Yahoo and Feedster

Today I was offline most of the day, as I was working on the book. I met with key folks at Yahoo Search and Scott Rafer , the reasonably new CEO of Feedster (Feedster is a search engine which trolls only RSS feeds. What’s an RSS Feed? Read this). Lots of discussion about the present, past and future of search. I left Yahoo impressed with where the company is and how the folks I spoke to approached the problem/opportunity of search. Yahoo has the longest history and the richest resource set of any of the current search players, and they certainly have got serious search religion, something that definitely waxed and waned over the past five or so years. Much talk of the lessons of Google, the looming competition with Microsoft, and good discourse on the best ways forward w/r/t search solutions. Ninj Srinivasan, employee #5 and Editor of Yahoo, has been there for nearly 9 years and is still stoked to come to work every day. That says a lot to me. Funny side note: while I was there, 1600 Yahoo employees and various friends went out into the parking lots and broke the world record for simultaneous yodeling. I’m not kidding…

Meanwhile, Scott at Feedster reminded me that PageRank was created back in a time when making links was pretty hard to do, and therefore a scarce resource, so one could reasonably trust that the links were authoritative. But, he supports the idea (as do others, see here and here) that blogs have muddied the waters for PageRank to the point of diminishing returns. I am certainly not one with the chops to judge, but I do wonder how this meme is playing out now that it’s been in the world for a while? It’s not like anyone has come up with anything demonstrably better….
In a related note, Scott mentioned an intriguing development at Feedster that he calls Feedpaper. From his site: “Feedpaper is a dynamic newspaper constructed from RSS feeds around a particular topic.” They are playing around with the idea on the site, for examples, check out this one on the Dean campaign and this one on the recent Longhorn developer conference (Longhorn is the new version of Microsoft’s interface, due sometime later this decade…). Scott’s thesis is that algorithmic search engines do a pretty poor job of aggregating this kind of content for readers interested in a very specific “island” of information. Watch this space, it could get interesting.