Mike makes a good point here in response to Tim’s point of view, and Tim counters here. It’s a very interesting debate, one between two folks I’ve partnered with for some time (Mike’s TechCrunch is part of FM, and Tim and I have partnered on any number of things, including Web 2, Make, and several sites that are also part of FM). So I’m a bit conflicted as I watch these guys debate.
However, one thing I will assert: Search is more than a subsystem of Web 2, as Tim puts it. I think search has become the interface to Web 2, and so I agree with Mike that Microsoft should not abandon it. It’s how we navigate the world of knowledge, and it’s way too early to say we’re done with the evolution of that navigation. But as Tim also points out: “a platform beats an application every time”. Damn right! That’s why I think Microsoft (or Yahoo) should innovate in search, as I pointed out here and here. If Google were to open its platform up, man, then it might well be game over. An open platform that has near monopoly share? Now that’s something to ponder.
i disagree that search is the interface of web20. search is an interface left over from web10, and in the absence of decent discovery interfaces/friend recommendations (we finally, FINALLY, have that in an efficient form with twitter) search will do.
but social recommendations are the interface / filter for effective web20, and that is the real deal.
soon as we do the thing we discussed the other day better (inference engine), search becomes 20% and discovery 80%. and discovery is the interface of web20 (as twitter has become for me).
i disagree that search is the interface of web20. search is an interface left over from web10, and in the absence of decent discovery interfaces/friend recommendations (we finally, FINALLY, have that in an efficient form with twitter) search will do.
but social recommendations are the interface / filter for effective web20, and that is the real deal.
soon as we do the thing we discussed the other day better (inference engine), search becomes 20% and discovery 80%. and discovery is the interface of web20 (as twitter has become for me).
Not to demean the significance of search but I think it is not fully 2.0 but maybe something in between the transition but a big part unarguably. As 1.0 is information -driven so was search a truly significant part of it.
Now that collaboration and community are big in 2.0, it’s time new tools emerge with twitter potentially waiting to be harnessed. Can search crawl into it. Not sure but I think many will try or have been trying already.
Best.
alain
I think I better go on record defining what I mean by “search” as it includes your point around social filtering, Mary, and Twitter, Friarminor.
I was about to jump in to the comments with a similar reaction, about how there is no real “web2.0” in search. It might be a very cool and useful thing, but that in itself does not make it “web2.0”. Defining everything cool and useful on the web as “web2.0” is a bit self-serving. But I see y’all have already covered the issue.
However, let me just point out that when John says that “search” includes “social filtering”, what I think John is really talking about is “information retrieval”, not “search”.
Information retrieval, for decades now (long before “web2.0”) has been working with the social aspect of information needs. On the social/collaborative recommendation side, there was an early paper at Xerox PARC (published in CACM, December 1992). On the search side, there was the work from the early 1970s from Eugene Garfield on citation (i.e. “hyperlink”) analysis, which is also an aspect of the social component.
But let me just clarify one more thing. John writes:
[Search is] how we navigate the world of knowledge, and it’s way too early to say we’re done with the evolution of that navigation.
I just want to point out that “search”, or “information retrieval”, is more than just navigation. It is even more than just evolved navigation. As I quoted here in this blog a week or two ago, Andrei Broder had that 2002 SIGIR Forum paper on the three types of “search”: navigational, transactional, and informational.
My big problem with most, if not all, of the major search engines is that their focus is on navigation, rather than on, say, information. They’re leaving whole swaths of “search” untouched.
So let me just expand on this statement, and say that it is not only way too early to say that we’re done with the evolution of navigation. But it is also way too early to say that we’re done with the evolution of information seeking, of any kind.
And speaking of Microsoft giving up on search, I think it’s worth pointing out that Google is also giving up on important, previously held ideals.
Case in point: I was recently at merriam-webster.com, doing a few dictionary searches. And I noticed the big huge graphicals ads on the right hand side, that would appear after each search. They were “Ads by Goooogle”.
Ok, so these animated graphical ads were kinda gaudy. Fine. Google has started saying recently that sometimes the best results are of the same genre, i.e. an image ad is the most relevant type of ad to an image search. Nevermind that I was doing a text search and not an image search, but whatever. I’ll accept that there are times when a graphical ad can be the most relevant.
So then I looked the content of the ads themselves. There was an ad for Nascar. A political ad about Obama. An ad for the new Toy Story attraction at Disneyland. Another political ad about Hillary and McCain. An ad for cars.com. An ad for an Orchard Bank MasterCard. And an ad telling me, as a new mother, to get vaccinated for whooping cough to protect my baby.
Let me note that my query, and thus the text on the resulting page, was for the word “dishwasher”.
These “Ads by Goooogle” were not only gaudy, they were also completely irrelevant. Or even sexist, if there was some sort of search engine intentionality behind the match between “dishwasher” and “new mother”.
To see Tim O’Reilly and Danny Sullivan argue that Google has won the war, when these ads are so completely irrelevant and therefore against every principle Google ever stood for tells me that the battle is far from over. And while I am not a big fan of either Microsoft OR Google, I think the last think that either company should do is “give up”. I don’t want Microsoft to give up on search in general. And I don’t want Google to continue giving up/selling out by showing gaudy and irrelevant Nascar, credit card, and disneyland ads.
John,
Don’t be confused by the monetization engine, and then ask yourself of your online time, how much is driven by search? If it’s the UI, it should be involved in most of your activity. For me, it’s a pretty small piece.
So I argue: it may be the most valuable subsystem right now, but that doesn’t make it the master key, especially as we move to mobile.
Whoever gets the identity subsystem right will have a very strong position, as will whomever dominates location, among others.
@Tim – I think I need to write a longer post about what I feel “search” really is these days. It’s for sure not just Google or Yahoo or Microsoft. I think it’s a metaphor for how we navigate the web in general.
Tim, could you explain what you mean by how much of your time is spent in the search UI? I don’t quite get what you mean.
Because when I think about searching, ALL of my time is spent in the UI. I’m either trying to formulate just the right quer(ies) in that little box (with little or no contextual UI assistance from Google, I might add) or else I am walking through a list of results, one at a time, top to bottom.
Everything about that entire process is UI. Why, for example, are results laid out in *rows* from top to bottom, so that I have to scroll down to see all 10? Why are results instead not laid out in *columns*, from left to right, so that I can see the top 10 results all in a glance? Then, if none of the ten meet my information need, I can scroll down, and see the ads below the fold, also laid out in columns.
The fact that the UI isn’t laid out in this manner is a conscious decision on the part of the search engine companies. And for every single search that you or I do, we are dealing with the existing UI. 100% of the time we are in that UI. It’s not just a small piece, it’s 100%.
But maybe I’ve not quite understood you. Could you explain what you mean?
@John: “I think it’s a metaphor for how we navigate the web in general.”
Again, let me take mild umbrage at you, for continuing to think of search only in terms of the “navigation” metaphor. Search is so much more than just navigation.
However, the big web search engines don’t currently actively support any other type of user behaviour other than navigation, so most users have come to think of search only as navigation.
But that is only a limitation of imagination (on the users’ part) and service (on the web companies’ part).
If you look at search’s historical roots, you’ll see that it is much more than navigation.
Information gathering and filtering, pattern discovery, topic detection and tracking, new event detection, information boundary exploration, information routing, question answering, document summarization, etc. These are all things that fall under the “search” umbrella. Not just navigation.
I agree, JG, that it’s more than a narrow definition of navigation. Search is the journey we take through information. I really do need to write that longer post.