For most brand advertisers, success is ultimately measured after the completion of an advertising campaign -- often much later. Depending on the nature of the product or service (for example, a movie premiere vs. a CPG product launch being at the two extremes), it can take time to measure market share gains, units sold and so on. Online video can be an interesting testing ground for brands due to the variety of interaction rates available, providing an early snapshot or preview on the branding effect.
The availability of interaction information is so compelling for marketers that there has been some movement toward pricing on a cost per interaction (CPI) model instead of the standard per-impression (CPM) pricing metric. Advertisers are obviously intrigued about the prospect of paying only for those users who have interacted with their ads.
This examination by advertisers has caused some stir in the industry on the publisher side. For small- and medium-sized publishers, offering video inventory on a CPI basis can be a great way to get on the radar for a marketer or their agency. They can "prove" the quality of their user base and ability to accommodate cutting-edge formats. Indeed, most CPI efforts have been focused on rollouts along the mid-to-long tail for this reason.
Where CPI will ultimately struggle is along that upper tail -- brand-name sites that attract a premium audience and the resultant premium pricing. Most feedback I've heard from brand-name publishers has been largely negative about testing CPI. These sites feel that the brand association provided by their site may not be accurately taken into account via a CPI pricing structure.
So where does CPI go from here? Is it destined to be a way to leverage the long tail? Will advertisers, with their wallets, force a sea change among their long-standing, premium partners to CPI pricing? Or will a totally new model emerge and become the way online video is bought, sold and evaluated? Time will tell.