Forget Doubleclick, the future is PPA
It amuses me to see the online advertising vendor DoubleClick up for sale, supposedly for $2 billion or more. Whoever buys DoubleClick at that price will not see a repayment on their investment. Therefore I surmise that Google will win the alleged bidding war with Microsoft to acquire the company, since Google has a track record of throwing large piles of cash at iffy acquisition targets. If Microsoft wins the deal, then we will know just how worried Microsoft has become about its ability to remain a player in the online advertising business. But buying DoubleClick will not resolve Microsoft's difficulties.
The most important asset DoubleClick possesses is its relationships with publishers and advertisers.DoubleClick's 2006 'advertising as a service' revenues equalled AdCenter's The problem for any buyer with a significant existing portfolio of websites is that if it tries to leverage the advertiser relationships by promoting its own properties to them, then it will harm all its publisher relationships — including AOL, DoubleClick's biggest publishing partner.
By the way, there are two reasons why the sale of DoubleClick is a big SaaS story. First of all there's the widely-held belief that advertising will become an important revenue source for SaaS vendors — and that's true, although not the sort of advertising that DoubleClick handles, which is a point I'll return to in a moment. Secondly, there's the rather less well-known fact that DoubleClick is itself a significant SaaS vendor. Although some of its customers license the company's DART ad serving software for installation in their own data centers, a substantial proportion — including AOL — use DART as a service, hosted and operated by DoubleClick. According to a source quoted in The Wall Street Journal (subscription required), and with my emphasis added:
"DoubleClick offers services for online advertisers, ad agencies and Web publishers for managing, delivering and measuring online advertising. The company had roughly $150 million in revenue last year, according to a person familiar with the matter. More than $100 million came from serving ads for publishers to their Web pages and delivering the ads to be served on behalf of advertisers."
See? No mention of selling software. DoubleClick is an advertising-as-a-service company, generating more than $100 million in annual revenues. That gives it quite a lot of synergy with Google and Microsoft of course. Not many people realize that contextual advertising is a multi-billion dollar software-as-a-service category. But regular readers may recall that last year in a posting called Google's profits under attack I pointed out that "AdWords is an on-demand application" and therefore Google itself is a "first-mover ... in delivering online advertising as a service."
According to some as yet unpublished research I've been doing recently, Microsoft's AdCenter has not so far made as much progress as I'd thought it might last May. In fact I estimate that its advertising-as-a-service revenues for all of 2006 (after stripping out the publisher's element attributable to Microsoft's own properties) come to about the same as DoubleClick's. So at least if it bought the company it could make AdCenter's numbers look a little more respectable.
But the real problem with DoubleClick is that it serves banner ads. That's so last century — indeed, Tim O'Reilly even put DoubleClick at the top of his list of Web 1.0 examples in his seminal 2005 essay What is Web 2.0.
The future of Web advertising is not even contextual advertising, as delivered by Google AdWords, Microsoft AdCenter or Yahoo! Panama. Indeed, for many transactions, the Web makes advertising redundant, because it can put what users are looking for (and willing to pay for) directly in front of them at the point where they need it. This is more like the affiliate model, in which website owners get a cut of revenues earned when a visitor performs some kind of action as a result of clicking on a link on their site.
Google moved towards this model last month with the release of its Pay Per Action (PPA) beta, which pays a fixed sum when visitors complete an action defined by the advertiser. It's cumbersome, though, as website owners must opt in to the service for each individual advertiser, and it's not served contextually.
As if by coincidence, the day after PPA arrived, Amazon announced its beta of a new service for affilates called Context Links. This moves Amazon's affiliate program firmly onto Google's turf, as it automatically serves ads for products based on an automated contextual analysis of the page content. But of course it's not quite an advert because site owners only get paid if the visitor actually buys something from Amazon — in other words, PPA (or more accurately, PPP — Pay Per Purchase). This is not the first time that Amazon has, in my view, outclassed Google in its approach to monetization. But oddly, the announcement went almost unremarked.