In TV, Do they Call Craig Ferguson Remnant?
November 2nd, 2009 by Kathryn Koegel
The Internet is a medium with infinite content sources and dispersed usage. Here’s the bad news: It has a really crappy way of valuing inventory. Ad networks have a part to play here. They make this broken media easier to use by aggregating inventory and actually making the medium more efficient to buy. What they have negatively produced is a layer of confusion to the marketplace with no one really sure who is representing what inventory.
Inventory is being sold and resold which results in deals that do not deliver or retain latency in the ads actually showing up on the page. On the positive side, in their quest to differentiate their offerings, ad networks introduced principles of data modeling long understood in the direct response world to online. After all, there is only so much contextual inventory to go around and since advertisers have to pay a premium to get it, ad networks created a value out of all the rest. They’re even responsible for some great new terminology around what they are doing: Some are calling their business “The Second Channel” or “Non-Guaranteed Inventory.” It’s better than “remnant” but not much.
Ad networks have put their principles to the test by creating elaborate profiles through their cookies – and since the publishers have for so long provided the sugar –gazumped them of their audiences. Questions abound about who owns a Wall Street Journal user once they are not on the Journal. But since that data is out there, the Journal can argue to the IAB task force on data ownership as long as they want with little result. The data exists and the networks that have that profile aren’t going to suddenly, completely and finitely turn it over. Publishers should be especially wary of all of the free tools for measurement that have been introduced over the past two years: read data usage policies carefully whenever adding a company’s cookies to help measure a site. Those cookies are veritable Trojan horses that can and are being used to build behavioral profiles that those companies are then selling themselves.
Premium publishers should get into the game themselves and model to their advertisers’ content, thus creating value for their non-contextual inventory. After all, they have audiences advertisers value, and can guarantee the transparency marketers seek. Many of the top publishers effectively use behavioral targeting tools like Platform A, Tacoda and Audience Science to do just this. Some media conglomerates are pooling all their “remnant” cross company and have started elite audience networks (e.g. Time Inc., Forbes, Martha Stewart – in October, a group of magazines announced their own network). There is no magic to building a network. With ad exchanges and products like Google’sNetwork Builder tool for publishers, anyone can pool and sell inventory.
Online is probably unique in having early on determined that what was most saleable like home pages and contextual placements was “premium” to be sold through direct sales forces, and everything else was “remnant”: Cinderella pre-makeover to be dumped on an ad network who would do the “bibbety boppity” thing they didn’t have time for. It’s often cited that 80% of online advertising is remnant (latest example, MediaPost 10/7/09 article on the Rubicon company). Google’s Q3 earnings report noted that a much more modest 25% of their network display went unsold. This practice – and even the nomenclature – has held the medium back.
Think of how inventory valuation works in other media. In TV the most premium is sponsorship with an ad package around an event: in effect flat rate inventory with a premium paid for association with something like the SuperBowl or the Academy Awards. Next comes primetime. Then day, late night, and fringe. Yes, the inventory is tiered for pricing purposes, but they’re selling the various values of these audiences. They don’t call Craig Ferguson: “Remnant.” The same holds true for print: advertisers pay premiums for certain positions and there are volume discounts but they are essentially buying one audience for each publication: there is very little “remnant” involved.
Now because networks and the behavioral targeting companies that also built networks have been so good at turning “remnant” into the hot chick who can make an appearance at the ball – they are in effect delivering to mass advertisers exactly what they have always wanted: an implied demographic bucket or contextual relevance on a broader scale at a better price.