Data Dashboard: The Call of the Trading Floor
There has been much speculation as to how, when, and to what extent the media industry will look and behave like just such a trading institution. The recent explosion of paid search marketing, with its Vickery-style system, has made some observers wonder: Will the future of media, with different forms going digital, be based on an auction system? Will we sit on trading floors, constantly placing bids based on expected return and risk? Will we switch from being dealmakers to arbitrageurs?
The concept of trading was born from the precept of investment optimization. In no other media space can one measure the flow of different investment levels to sales return as quickly as with paid search. An agency can optimize in and out of keywords as fast as a bank can buy and sell stock. The first barrier to media trading within and across channels will be establishing the platforms by which to do it. We'll need to create for print, TV, radio, and display ads the infrastructure that's already commonplace for search.
In October 2004, Chris Anderson, editor-in-chief of Wired magazine, published an article entitled "The Long Tail," which described how the Internet has thrown Pareto's 80/20 rule out the window and allowed companies to profit from the "long tail" of their offerings. In the media planning world, it has meant a fundamental switch from buying a few big properties to efficiently buying hundreds of small ones, as well. The latter will be the driving force for media trading as a dynamic bidding system.
The key to a successful bidding system is being able to move in and out of the market quickly. The major search providers are well positioned for this because they built their platforms with both speed and extensibility in mind. Google has constructed a formidable and eclectic network of Web publishers to serve contextually targeted "Ads By Google"-sponsored links. Since last year, Google has allowed advertisers to bid on keywords that serve display advertising to partner sites, bypassing traditional display brokers and ad networks. The appeal of this long-tail platform drove 42 percent of Google's Q4 2005 revenue.
The ad networks, especially those that create behavioral targets, like Tacoda Systems or Revenue Science, will likely evolve along these same lines. As they build out hundreds if not thousands of behavioral segments across myriad sites, media agencies will manage the segments like they would a keyword list, perhaps establishing a Vickery auction for behaviors the way Google's AdWords does for contextual links.
The offline world is also beginning to experiment with the long tail. For example, Spot Runner, a Los Angeles-based startup, gives small businesses the ability to advertise on local TV by making commercial production, media planning, and media buying services fast, easy, and affordable. Google's acquisition dMarc Broadcasting connects advertisers directly to radio stations through its automated advertising platform.
The extent to which the media industry adopts bid-based systems will depend on more than just technology. Content owners will have to determine what placements they'll be willing to open up to auction. Yahoo, for example, might never release control of its home page placements. Those positions might always be sold in conference rooms with a handshake. But they may eventually conclude that a text link at the bottom of a MyYahoo page could be more profitably sold at auction.
Media agencies need to start thinking and acting like financial institutions. To do so, the agencies will have to make substantial investments in data systems that log media activity and conversion fulfillment minute by minute. They'll also need to hire talent capable of applying portfolio theory to the media marketplace by tinkering with tried and true formulas of valuation.
Adam Ghahramani is media strategist and Greg Rogers is director, strategy & insights, for MEC Interaction, the online, search, and direct response business unit of Mediaedge:CIA. (email@example.com and firstname.lastname@example.org)