This week, Chris Bowers at OpenLeft has been encouraging readers to run their own media campaign. The idea is very simple: at a fairly low budget, anyone can set up a simple Google ad campaign, targeted geographically and by keyword. Bowers has been running two ads - one against McCain, the other against Palin - in his native Pennsylvania, and thinks he can reach a lot of voters on a fairly low budget.

The commenters at OpenLeft have been ecstatic about the idea, and I think it is exceptionally clever. In addition to the first order effects - exposing anti-McCain messages to a lot of voters in swing states - the campaign could also have an indirect sway over the campaign's own messaging strategy, by demonstrating in a quantifiable way the messages that work (and receive a lot of click-throughs) and those that don't. I suppose that's a long shot with this campaign, but it's nevertheless a possibility.

In any case, I'd be interested to see if someone could take this idea to the next level, and make the decentralized media campaign idea a bit more social.  For example, would it be possible to set up a website which allows people to set up all of the parameters for a Google Ad campaign - the keywords, the geographic target, and the message/link which appears - and then to aggregate all of those campaigns on the website in some interesting way?  There are a lot of different ways to do this - e.g. breaking down ad campaigns by state, tag-clouding the chosen keywords, and showing aggregate click-through and impression statistics.  This kind of aggregation could be augmented with comments (suggesting refinments and tweaks to existing campaigns) as well as team fundraising pages, allowing site visitors to support one campaign or another monetarily.  It's also possible to maximize and quantify the impact of a campaign like this by targeting all of these ads at an action-oriented microsite, which takes a user through the steps of signing up for Obama's email list, giving a small donation to the campaign, signing up for My.BarackObama.com, and so on.

This sounds, to me, like a good example of a simple business idea that could be modestly profitable, since after all the main point of this project is to sell Google ads.  Technologically, this should be a relatively simple mashup of a community platform like Drupal with the Google Adwords software development toolkit.  With relatively low costs, it should be possible to set the transaction costs - on top of the raw costs for the campaign itself - at a sufficient level to generate relatively decent profit margins.  Besides the constant problem facing any social web platform - will anyone show up? - the only difficulty, as many OpenLeft commenters have already alluded to, is whether such an endeavor would run afoul of campaign finance rules, and whether or not Google Ad purchases would be considered campaign contributions.  My guess is that this kind of project would have to be organized as a 527, or under the auspices of one.

If something like this doesn't take shape between now and Election Day, it's probably a worthwhile organization to develop, even so.  Beyond the immediate need to go on the offense against McCain, as Chris points out, it's important to help develop and test messaging for Democratic candidates up and down the ballot, in an environment that's not controlld by the campaigns themselves.  More than that, with Google beginning to sell offline media ads in newspapers, radio and television, there's no reason to restrict an ad campaign to Google Adwords (although it's much more gratifying, as real-time metrics are available.)  Any takers?

Disclosure: My company worked on a small technical/design project for Open Left last year.