Dead Again: Amazon’s A9 and the Future of Search

In an otherwise irritating AP article that sweatily tries to make the case for Amazon as Big Brother, there is this shit-disturbing quote:

Analyst Mark Mahaney with American Technology Research also questions whether A9 is worth the hefty investment.

A9 ranked 41st in popularity among search engines in February, according to Nielsen/Net Ratings, attracting only a fraction of visitors to Google or Yahoo.

“It’s a little more of a stretch to me as to why investors should be excited about that,” Mahaney said. “Isn’t that a little crowded space?”

Now, to be fair, many people would have said the same thing about Google back in 1998 — and many did say as much, with various VCs having conceded that they would have or did pass on the fledgling search company. After all, so the 1998 wisdom went, Search was soo-oooo 1995.

Is analyst Mahaney falling into the same trap? Good question. He plainly is if he is saying that A9 is 41st and, therefore, it is nowhere in a crowded space. Google was almost certainly 41st early in its history, and things turned out okay for Sergey and Larry (and Andy B.).

But another way of asking the question is to ask whether A9 brings as much to the table, relatively speaking, in 2005 as Google did in 1998. Back then Google was plainly better at delivering on-target search results, and that is why I (and many others) noticed the difference immediately and made the switch. Such is not the case with A9, in that it is not obvious to me that its results are any better than Google’s, so the same comparison isn’t driving a speedy change of search tools.

Isn’t it possible, however, for there to be other bases of comparison? Sure. For example, many would argue (as I did with a friend at lunch yesterday) that Google currently wins largely because of its interface. Yes, the search quality is good, but that has become reasonably solid across most of the major search engines. The main reason why I stick with Google is the main reason I stick with most software apps: Because I know how its interface works and I’m comfortable and productive using it.

And A9 is trying to do something different. Rather than competing on a me-too interface and better results, it has been trying to compete on a radically different (better?) interface and (sort of) me-too search results. While that hasn’t yet been a big winner, it is also a slower sell: People get comfortable, and there are significant switching costs to changing interfaces, no different than typing in a Dvorak layout versus using Qwerty (okay, bad analogy). 

I’m not convinced that A9 will win with this approach — I find its current interface busy and difficult to sort out — but that is not the same thing as agreeing with Mark Mahaney and saying that search is dead (again).


  1. a9 does some cute work with dhtml (sliding columns etc), but in terms of actual results, i don’t find the amazon data much of a value add.

  2. If/when amazon makes A9 search the text of books along with the regular web search, then i think that would change the game.