Amazon's Search Portal

Yeah, you read that right. Amazon's got a search portal (well, they've got a wholly-owned subsidiary that's a search portal.) Pardon me whilst I go get an aspirin. They're not running their own crawler; they've got a strange mishmash of Google, Alexa, and Amazon results. But I think this search engine launch is very, very important. I'll explain as we go. First go to http://www.a9.com .

The search box looks very basic. Enter a search. Once you do you'll get Web results that look like Google results, except for the "site info" button. The "site info" button generates results from Alexa which include screen shot, page rank, and other information. (Those pages are hosted on Amazon.) (This site also offers a toolbar, which offers a pop-up blocker, Web search, and so on. Since the toolbar *requires* an Amazon account, according to the requirements at http://toolbar.a9.com/ , I am less than enamored and I'm not going to cover it deeply here.)

If you look over on the right you'll see that there are other tabs. One of them are book search results, which of course provide results from Amazon. If you have an Amazon account, and you sign in, you can also review your search history. When I looked at it of course there were only searches for the last 24 hours. You can edit the searches here. All three search boxes are spread out on one page of results horizontally and for a wonder I don't have any complaints about the layout. The font size is fine and I can actually see everything.

If you have an Amazon account you may be a little hesitant to log in and have your account attached to your search results. You do not have to log in to get search results. Furthermore, Amazon also provides a site -- http://generic.a9.com/ -- which doesn't use personalized features and doesn't even set cookies. That's a nice touch. Another nice touch is that you can run a query from your browser status bar, just by typing http://www.a9.com/query. I tried http://www.a9.com/"three blind mice" and A9 had no problem.

I was talking about this with someone else and they were less than enthused. "It's just a Google search. Who cares?" While I agree that I would have been much more impressed if A9 had had its own search index, I still think this is a giant step forward. This is an integration of a discrete set of content (book results, and content from books) that in toto are more credible than the Web as a whole. (I said in toto. I'm not going to argue the credibility of individual books.)

It is my opinion that we are going to see this more and more often. The GILS conference smacked me across the head with it; there are simply too many valuable collections of data that are being ignored or insufficiently indexed for a variety of reasons. In a perfect world, of the site search companies like Atomz or FreeFind would team up with Google. They'd seek out these data collections, offer to index them, then hook up their indexes to Google's main index. Then Google and these partners could split the ad revenue on search results that come from those pages.

A9 is still in beta. I will be watching this closely.

Posted to Search Engines-Various | TrackBack