Via John Battelle, the latest entry in comScore's take on Internet search market share. This tends to be one of the conservative takes re Google's share, and has a decent time series behind it, which I've knocked into a quick and dirty graph:
(For simplicity, I've omitted the Ask Network, which has been essentially flat at around 5 percent during the same period, and the AOL/TW complex, which has dived from 9ish percent to about five.)
Other than Google's kicking everyone, what conclusions might be drawn from this record? One is the collapse of the notion that users' means of access would control where they spend their time and do their searches on the net. Whether your strategy hoped to influence the outcome via control of the user's browser and PC (MSFT), dial-up (AOL) or broadband access (YHOO, via AT&T partnership), this chart shows you losing. Anyone just hitting the nets now is a self-proclaimed late adopter. Everyone else, from the pioneers to the late majority, have been out there long enough to find their own way around. The 'control the user' notion was probably bogus from the start, but it's visibly dead now. (Wireless providers might profit from the lesson.)
Another conclusion is how a market has been beating a traditional sales model. Advertising is of course the payload of the 'search' marketplace. From the start, Google used a open, electronically mediated, bid market for placement on its pages. Yahoo went and bought what it thought was its own search ad market, found it that it had acquired a 'fixer-upper' and then blundered through the rebuild. The better part of its revenue came from traditional banner ads, sold with a traditional sales force and it is only now nominally catching up with its Panama release. The battle has spread to placement on other net properties, with the dual acquisitions of DoubleClick and Right Media. Terry Semel uses the latter occasion to slap at Google:
We think supply and demand should be regulated by the marketplace, not a closed platform.... We think our open approach is a clear differentiator from others in the industry and will provide significant benefits to publishers and advertisers.
In reality, Yahoo is coming to this point of view fairly late in the game, and that's been part of their problem. There's no inherent need or desire on the part of advertisers to have multiple markets to which they must go to get out their message. Provided the operators of the markets don't try to extract too-high monopoly rents, there's a lot of reason to desire a single venue with all buyers and sellers present and transaction costs driven to a minimum. Critical mass matters a lot in that case. Revisit the chart above. If the value of a network rises as the square of its size, how far is Yahoo actually behind?