Reading a New Yorker article on behavioral economics (via Paul Kedrosky) got me thinking: what if the division of ad inventory into premium (sold in advance, guaranteed, generally for higher CPMs) vs. remnant (sold real-time through ad networks or exchanged, non-guaranteed, generally for lower CPMs) is causing a form of anchoring?
In other words, is ‘remnant’ inventory cheap simply because we’ve all been telling ourselves over and over again that remnant inventory is cheap? Reports of rock bottom CPMs act as the anchor for the initial estimate. Advanced targeting and optimization boost up what people are willing to pay from the anchor. But because the anchor’s so low, advertisers are never willing to pay that much, to the point where it’d compete with what they’re willing to pay for the premium (which equals ‘expensive’) stuff.
I’d love to see an experiment where a major publisher said ‘hey, we’re canceling our premium ad sales – from now on there’s no premium or remnant, there’s just inventory.’ In other words, they’d sell the ‘premium’ stuff using ‘remnant’ methods. Would the inclusion of ‘premium’ in the inventory pool be enough to remove the ‘remnant = cheap’ anchor, and boost CPM rates overall to the point where the increase would compensate for the loss of the premium? Or would the use of ‘remnant’ buying methods still keep the anchor in place and the prices low?
One possibility, which would be hilarious: the ‘remnant = cheap’ anchor is triggered by the buying method, not the actual ‘guaranteed vs. non-guaranteed’ differentiator. If you increased the annoyance of buying remnant and started forcing clients to fax paper contracts back and forth via fax – anything to get that ‘premium feel’ – would they be willing to pay more?