This image was lost some time after publication.

Is it wrong to want Google to win Doubleclick? The Wall Street Journal reports a rumor, probably put about by bankers drumming up excitement for the online advertising company, that the Mountain View search engine has entered the bidding. If Google was to win, it would be the search giant's biggest acquisition, probably worth more than $2bn, more than was paid for Youtube. The intervention of Google would be bad news for Microsoft, which was being touted last week as the most likely acquirer; and a Google victory would trigger another round of wailing at the search engine's incipient monopoly of online advertising. But there's a reason why monopolies develop: they're often convenient; and the absorption of Doubleclick by Google would fall into that category.

The biggest problem that online publishers face is that there's no good way to sell their "remnant" inventory, the ad impressions which aren't sold directly. Even on sites desirable to advertisers, which are sold out, there's often plenty of space, because its difficult for site owners to project traffic levels: advertising is sold days or weeks before it runs; and a popular story, and an unexpected burst of traffic, brings no immediate benefit. In categories such as political news, the bulk of a site's inventory can remain unsold.

Existing ad networks — which seek to match up last-minute advertisers with spare inventory as if they were pricing standby tickets on an airline — generally do a poor job. Remnant impressions typically sell for a tenth or so of a slot's list price. The only network which has worked for publishers, and advertisers, is Google Adsense — but the search engine's contextual text ads require their own dedicated space, so they don't solve the problem of unsold leaderboards, or skyscrapers. Google's marketing of these display ads has not been successful.


So, why would Google's acquisition of Doubleclick make a difference? The New York advertising group runs the platform, DART, which many publishers use to host the ads they sell directly. It's a system which allows ad sales departments to set the order in which different campaigns run, cap the frequency with which an ad is shown, and squeeze more revenue from a particular slot.

I don't think this has been reported yet, but Doubleclick is launching its own ad network, which would give publishers the option of running default ads, supplied by Doubleclick's partners, when no direct campaigns were running. This would be an improvement on current arrangements, which require, when a user hits a page, a call to the Doubleclick server to check whether there are any direct campaigns, then another call to a third-party remnant network such as Tribal Fusion and, if no remnant campaign is available, a third call to pull up house ads. Messy.


But imagine if Doubleclick's new network, to which most of its publishing clients would subscribe, could tap Google's vast array of advertisers. Google, with Doubleclick, might be able to achieve critical mass in banner advertising: collecting together enough publishers and advertisers for a liquid 'spot' market to develop, as it has in text advertising. And that would be great news for online publishers. Someone else can worry about Google's monopoly.