My prediction is that A9 will quickly introduce “collaborative search”

My prediction is that A9 will quickly introduce “collaborative search”: As much as I admire the insight and analysis of search engine blog guru John Battelle, I think he’s missing the obvious in his inside scoop on the launch of Amazon‘s new search engine, A9.

a9I highly recommend that you read what John Battelle says about A9 and play with it. It is awesome. However, I think Battelle’s whole Google vs. Amazon riff is a red herring. To me, A9 is not designed as an Internet search engine, but as a knowledge-searching tool to end all knowledge searching tools. Jeff Bezos is a brilliant person, but he doesn’t invest in things that don’t have (at least in theory) an understandable business model. And nothing in his history displays a belief that advertising is an understandable business model.

I don’t think Amazon wants to compete with Google. Google admitted recently that it was a content business. Amazon has no such designs. Amazon, rather, wants to connect you with something you can purchase. And it has learned like no other how to power such connections by using the comparative patterns displayed by its customers. Any Amazon customer can tell you how they have bought a book based on Amazon’s revelation that, “people like you” have bought it.

I predict (and I am about to explain why I have at least some standing on this opinion) Amazon’s amazing skills at such application of collaborative filtering in providing search results will be A9s’ true secret sauce. As you look for information, Amazon will provide you the results that “people like you” have found most helpful when searching for the same information, product, place, answer, etc. With A9, however, I predict the results won’t be limited to URLs, but will link to passages in books, products in stores, scenes in movies, insights from reviewers (and bloggers?)…all based on the preferences displayed by people who have similar patterns of purchasing (and searching and clicking) as you.

Why do I predict this?

As two of rexblog’s five regular readers know, this is a topic about which I know, well, too much. Read the second question on the FAQ (written about four years ago) of the archived site, and you will see that I and several others made a big investment of time, talent and resources in pursuit of the belief that knowledge is a grassroots conversation and that search personalized via collaborative filtering algorithms developed by folks at the University of Minnesota and later at Net Perceptions and other places would be a great approach to matching up people with answers (or recommendations) to those with questions. (Some of the same ancient algorithmic DNA that helped power Amazon’s early recommendation approach.) It worked great for us…but then there was that that whole bust and all.

If A9 incorporates the collaborative filtering algorithms that power Amazon’s predictive recommendations to customers, it will (and I know this from very, very expensive first-hand experience) produce search results that will astound the user. Just think about it: Your search results will be filtered first by Google algorithms and then through Amazon’s collaborative filtering algorithms. In the simplistic metaphor we used at, “Your search results will be based on those results found most helpful by people like you.”

It will be cool. Promise.