As part of the weekly speaker series that our Technology Development Group runs (well, really Chad does all the work), John Battelle visited to speak of a room packed full of Yahoos.
He split the time between reading from his book and talking about his experiences researching it, his time at The Industry Standard, Wired, and so on. As expected, John was entertaining and interesting. He's really enjoyable to listen to. And having spent so much time digging into search for his book and on his blog, he has a unique perspective on this industry that many of us are living in.
There's an idea that he threatened to blog before I do, so I'll let him do that. I'll only mention that it's a search related doom-and-gloom screenplay. :-)
He also spent some time answering questions ranging from privacy to publishing and his vision for the future of search.
One of the things I like about John is his ability to stay down to earth, funny, and self-deprecating while also being one of the most visible figures in this industry.
John, sorry for the picture. I saw several folks taking pictures today, but Chad's blurry camera phone shot is the only one I can find on Flickr right now. And thanks again for dropping by Yahoo for the afternoon.
Posted by jzawodn at December 02, 2005 08:21 PM
I really enjoyed "The Search" as a Google success story.
Seems to me he didn't give Yahoo enough credit for it's early and brilliant insights into how "organizing information" would unlock the internet for mainstream users and how a corporate culture could succeed and thrive in a freewheeling, suit free environment. Didn't anybody ask about the cursory treatment of Yahoo in the book? Do they serve Google Kool Aid even at Yahoo?
Joe, he did a good job of explaining that right upfront.
What is hard to understand is why did Yahoo not put together its own search engine. The directory is a great asset and very relevant, but considering Yahoo's status and power, why did Yahoo allow AltaVista, Inktomi and Google to default to the organic listings after the directory listings on serps were exhausted.
They waited many years to buy a search engine and develop it - but interestingly enough, unlike Google and MSN they never started one from scratch.
Also why did they chose to buy Inktomi - were there any discussions with ASK / Teoma / Wisenut ???
Inktomi's SERPs just was nothing spectacular - there was no technological idea innovation - like the others.
Yahoo turned down Ebay, were they also approached initally by Page and Brin - when they were shopping for companies to buy their Search Engine Software???
oh - the fickle finger of fate :?
SEW - you should read the book - it's excellent. A key point was that very few understood the "value of pure search". The early search places wanted to become sticky portals and keep people within their own sites.
Search Engines Web ---
You ask a couple of questions: 1) Why didn't Yahoo create its own algorithmic search engine early on, and 2) Once they decided to buy not build, why did they buy Inktomi?
I can't answer the first, but I came to Yahoo! Search via the Inktomi acquisition, so I can answer the second to some extent. What I understood at the time was that Y! did a very careful eval of available algorithmic search engines (including evaluation of relevance) and Inktomi was the best according to those studies.
to SEW
in the early years of yahoo you could put your arms around the entire useful web. therefore you could actually build a directory that was of value. and we did.
as the web grew, this obviously became untenable. but for a time it was the best way to find sites. i know that sounds ridiculous now, but remember in 1996-98, users were less sophisticated and preferred things laid out for them.
obvioulsy things are different now.
>> SEW - you should read the book - it's excellent. A key point was that very few understood the "value of pure search".
this is not entirely true. that just plays into this ridiculous idea that the google people are otherworldly geniuses. in 1995 the market was not interested in deep mining, it was interested in making the web friendly. there wasn't even enough data to make a deep mining approach worthwhile. why buy 5000 machines to catalog 100k "real" websites?
battelle was obviously using the web at this time so he must realize this too.
timing matters. if yahoo came out in 1993 or 1997 instead of 1995 it would have also gone nowhere. you can be too early or too late even with a great idea. look at desktop.com. a full "ajax" site with apis exposed to developers in 1997. and now everyone is doing it. just ahead of its time.
Good points GrumpY (why do I think there are some great stories behind your handle dude?)
Timing .... is .... everything. After Google settles to earth and MSN catches up with Y and G I anticipate some great innovations in search from you otherworldly genius types, or perhaps even from ....outsiders....
I find myself mostly agreeing with GrumpY! once again.
Another bit of historical context is that, even later on (say, 1999-2000), search didn't look very interesting to business types. (This is an issue that Battelle does a good job with in the book.) It was viewed as a commodity service -- something a portal had to have (or license), but they couldn't figure out how to make money on it. (This, at any rate, was the view at Excite@Home when I was there.) The _really_ exciting stuff -- well, that was 3D interfaces, and lots and lots more micro-properties like astrology, etc. That's where we needed to invest.
Also, search was viewed as something that wasn't expected to need any more technical innovation. I mean, you found pages that had the words you were looking for - what else do you want? (Having been through this once makes me sceptical of GrumpY!'s view (from another thread) that search is done, and the battles to invest in are elsewhere.)
>> The _really_ exciting stuff -- well, that was 3D interfaces
strange but true, yahoo once had a vrml site for cruising the directory.
>> Also, search was viewed as something that wasn't expected to need any more technical innovation.
a couple of years ago i attended a seminar at overture research where a guest lecturer on recommendation systems opined that until google contributed link analysis (pagerank), the field had basically done nothing new for some time. interesting.
i still believe search is "done", but of course requires heavy (but fixed) maintainence. in my opinion the next step on the web is entertainment. compare time spent by the average world of warcraft user (hours per night) to the average google search user (minutes per week). no contest!! its a riskier business of course. i'm surprised the major web companies haven't moved on MMORGs in a serious way yet.
